The *COMPLETE* list of 27" 5K Displays?

[motrek]

Think I saw the info about PDF rendering on ycombinator, which tends to be people well versed in technical aspects of OS X. I don't think it is about vector vs bitmap, as normal desktop text does become slightly softer on certain non pixel-doubled modes.

Easiest thing is to test this empirically. Put your 5K into a fractional scaling mode in which you can see a difference in text/desktop elements. Then open a PDF and compare that to the exact pixel-doubled mode. I can never see any difference.

It sucks taking pictures of monitors, but the left-hand image is at pixel-doubled resolution and the right-hand image is at a scaled resolution.

The difference is subtle but I think it's clear that the left-hand image is a tiny bit sharper. So I don't think PDF rendering gets any special treatment in MacOS.

Also, I think it speaks highly of Apple and MacOS, if you have to compare side-by-side images with a microscope to tell if you're using a scaled resolution, and even then, the difference isn't immediately obvious.

(Edit: I deleted the attached image. There's a much better one in my post below.)

 

[EugW]

Also, I think it speaks highly of Apple and MacOS, if you have to compare side-by-side images with a microscope to tell if you're using a scaled resolution, and even then, the difference isn't immediately obvious.

The left is crisper but jaggy. The right is smoother but less crisp and a bit greyer in appearance. However, the image suffers from compression artifacts too which may affect the comparison.

Anyhow, the right picture is very similar to what I see with my 163 ppi Huawei screen at a non-2X scaled resolution. It's a 28.2" 3840x2560 screen, which is the identical pixel density as 27" 4K. I run at 2304x1536 (instead of the 2X scaled 2560x1707).

Anyhow, I'm still salty about Apple killing off sub pixel rendering. It makes the purchases of monitors that much more complicated, and expensive.

 

[motrek]

The left is crisper but jaggy. The right is smoother but less crisp and a bit greyer in appearance. However, the image suffers from compression artifacts too which may affect the comparison.
...

Good point about compression artifacts. I redid this experiment using RAW photos. Not only did that eliminate the compression artifacts, it also got rid of the moire, which I guess was being caused by Apple's photo stacking algorithms.

I'm hard-pressed to tell any difference in sharpness between the left and right images. If somebody told me that the image at the right was being rendered at native physical resolution, I might believe it, except that every time I do this experiment, the asterisk (V-naught *) looks very different at the scaled resolution, so I'm sure it's being rendered at a higher resolution and scaled down. (Ironically, the resulting scaled-down asterisk looks much sharper to me, not blurrier.)

 

[motrek]

...
Anyhow, I'm still salty about Apple killing off sub pixel rendering. It makes the purchases of monitors that much more complicated, and expensive.

Apple presumably killed it because it's incompatible with how they do scaled resolutions, which IMO is a pretty good reason.

(It's incompatible because sub-pixel rendering requires control over individual sub-pixels by strategically changing the colors of specific physical pixels. With scaled resolutions, the OS is rendering to a virtual resolution with virtual pixels, and the resulting image gets scaled down such that the virtual pixels don't correspond 1:1 with physical pixels. So the OS has no way to control sub-pixel brightness. If it tried to do sub-pixel rendering at the virtual resolution, by strategically adding colored fringes to text, the result would look like a multi-colored disaster when displayed on the physical monitor.)

The only time the lack of sub-pixel rendering would really matter is if you're running non-HiDPI on a low-resolution monitor... is that what you're doing, or intend to do?

 

[kvic]

Think I saw the info about PDF rendering on ycombinator, which tends to be people well versed in technical aspects of OS X. I don't think it is about vector vs bitmap, as normal desktop text does become slightly softer on certain non pixel-doubled modes.

Easiest thing is to test this empirically. Put your 5K into a fractional scaling mode in which you can see a difference in text/desktop elements. Then open a PDF and compare that to the exact pixel-doubled mode. I can never see any difference.

Do you happen to have the link to hacker news discussion?

An application can do its own drawing. But I highly doubt that it has the chance to do the drawing after the "fractional scaling" is already done.

 

[EugW]

Apple presumably killed it because it's incompatible with how they do scaled resolutions, which IMO is a pretty good reason.

Yeah but they killed it for non-scaled resolutions too.

The only time the lack of sub-pixel rendering would really matter is if you're running non-HiDPI on a low-resolution monitor... is that what you're doing, or intend to do?

It matters on any monitor less than about 170 ppi, HiDPI resolution or not.

163 ppi monitors vastly outnumber 218 ppi monitors out there.

 

[motrek]

It matters on any monitor less than about 170 ppi, HiDPI resolution or not.

PPI is irrelevant without also specifying a viewing distance. A 27" 4K monitor is 163 PPI (i.e., less than 170), but from 21 inches, it's 60 pixels per visual degree. Which is retina resolution. Meaning that a person with "normal" eyesight is going to be unable to resolve individual pixels. If a person can't even resolve ENTIRE pixels from that distance, what possible difference could SUB-pixel rendering possibly make?

And whether or not you're running HiDPI matters a lot. Let's say that "normal" text size is about 12 points. If you're not running HiDPI, the features of 12-point text rarely cover an entire pixel. Meaning that the ratio of antialiased pixels to purely black pixels is huge. So the number of pixels that could be potentially affected/improved by sub-pixel antialiasing is huge. But as soon as you turn on HiDPI, a lot of those pixels just become black. So the ratio of pixels that could even be affected by sub-pixel antialiasing is much lower, and thus, sub-pixel antialiasing overall makes much less of a difference.

 

[EugW]

PPI is irrelevant without also specifying a viewing distance. A 27" 4K monitor is 163 PPI (i.e., less than 170), but from 21 inches, it's 60 pixels per visual degree. Which is retina resolution. Meaning that a person with "normal" eyesight is going to be unable to resolve individual pixels. If a person can't even resolve ENTIRE pixels from that distance, what possible difference could SUB-pixel rendering possibly make?

163 ppi matters because as you say 21 inches is "Retina" for it... but recommended ergonomic guidelines are that a proper seating distance is 20" or more for a desktop. However, someone with better than 20/20 vision or vision corrected to slightly better than 20/20 vision can notice a difference even at 21-22". Actually, I have bad eyesight, but it has been corrected with my current glasses to around 20/20 or slightly worse, and it's still fairly easy to tell the difference. 163 ppi is very decent, but it's still noticeably worse than 218 ppi even at 21-22" (my normal seating distance). How do I know? I have both a 218 ppi screen and a 163 ppi screen on the same desk.

It's no surprise that Apple chooses a much higher pixel density than 163 ppi.

The other thing is I don't really like Apple's default font sizing for 2X scaled resolutions on 218 ppi monitors. My old eyes would prefer closer to 200 ppi for somewhat larger text at 2X scaled resolutions. Ironically, Apple sold 101 ppi monitors to pair with macOS for many, many years, but for some reason when they went Retina, they eventually settled on 218 ppi for desktops instead of ~200 ppi. I'd much prefer something like a 29.5" (30" class) 5K 5120x2880 16:9 monitor (199 ppi) or else a 30" 5K 5120x3200 16:10 monitor (201 ppi).

 

[tornado99]

Do you happen to have the link to hacker news discussion?

An application can do its own drawing. But I highly doubt that it has the chance to do the drawing after the "fractional scaling" is already done.

still can't find the thread, but one thing to bear in mind is that the framebuffer essentially is a pdf. Apple use the exact same engine to render PDFs and the screen itself. So splicing in a PDF to the post-rendered desktop is believable.

 

[tornado99]

Good point about compression artifacts. I redid this experiment using RAW photos. Not only did that eliminate the compression artifacts, it also got rid of the moire, which I guess was being caused by Apple's photo stacking algorithms.

I'm hard-pressed to tell any difference in sharpness between the left and right images. If somebody told me that the image at the right was being rendered at native physical resolution, I might believe it, except that every time I do this experiment, the asterisk (V-naught *) looks very different at the scaled resolution, so I'm sure it's being rendered at a higher resolution and scaled down. (Ironically, the resulting scaled-down asterisk looks much sharper to me, not blurrier.)

View attachment 2232107

I can see from the size of the pixels behind that your camera is closer to the screen in the left picture.

Anyway, at least to my eyes I never see perceptible differences in PDFs with fractional scaling vs pixel doubled. Whereas there is a very slight softening of other text e.g. web browsers (but not enough to make me go back to pixel doubled!).

 

[motrek]

still can't find the thread, but one thing to bear in mind is that the framebuffer essentially is a pdf. Apple use the exact same engine to render PDFs and the screen itself. So splicing in a PDF to the post-rendered desktop is believable.

Display PostScript - Wikipedia

en.wikipedia.org

Check out the section called "Modern derivatives."

Quartz 2D commands are executed to create rasterized backing stores that the Quartz Compositor draws to the screen:

Quartz Compositor - Wikipedia

en.wikipedia.org

My understanding is that all the backing store bitmaps are all the same resolution. This is the first I've heard that something might be rasterized to the monitor's physical resolution and not the user-selected virtual resolution. I'm still skeptical.

 

[motrek]

I can see from the size of the pixels behind that your camera is closer to the screen in the left picture.
...

I can see a grid in the images but I doubt those are pixels, probably a visual artifact of the camera sensor pixels not lining up with the monitor's pixels.

Note that the grid is at an angle vs. the text. That wouldn't happen if it was a grid of pixels.

 

[motrek]

...
Anyway, at least to my eyes I never see perceptible differences in PDFs with fractional scaling vs pixel doubled. Whereas there is a very slight softening of other text e.g. web browsers (but not enough to make me go back to pixel doubled!).

I have a 27" 4K monitor that I run at a virtual resolution of 1440p... text also looks soft to me (compared to a 5K display) when I lean in and look at it closely, but I don't know if that's because it's downsampled or if it's simply because a 4K monitor has less resolution than a 5K monitor. I assume the latter.

(Of course text on a higher-resolution monitor will look sharper. It's higher-resolution.)

I posted a comparison of native text rendering vs. scaled here:

External display 27"/4k - Scaling problem, really?

 

[kvic]

still can't find the thread, but one thing to bear in mind is that the framebuffer essentially is a pdf. Apple use the exact same engine to render PDFs and the screen itself. So splicing in a PDF to the post-rendered desktop is believable.

I got the sense you're confused. So perhaps you misunderstood whatever discussion happened on Hacker News that you read. Now about the new claim you just said above, you actually might be referring to intermediate buffers along the rendering pipeline for the desktop or GUI system. A framebuffer is already rasterised pixels, right before sending to the display. So it's not sensical to me to say a framebuffer in PDF format. Whether or not MacOS allows an application to draw directly to the framebuffer remains a question. I think technically it's possible. Any applications doing that? Highly doubtful, Apple's own apps included.

Easiest thing is to test this empirically. Put your 5K into a fractional scaling mode in which you can see a difference in text/desktop elements. Then open a PDF and compare that to the exact pixel-doubled mode. I can never see any difference.

Btw, this is not a proper experiment to verify your claim. It's comparing Apple to Orange.

A way to verify your claim will require custom apps. The closest tests I can think of is to set your 5K display in fractional scaled resolution. Open Microsoft Word or another word processor. Type some text. Generate a PDF. Open the PDF in Preview. Zoom in/out the Pdf to give similar text size to the Word. Now start pixel sniffing and see if you spot a difference between the two. You may also need to disable certain post-processing in PDF viewer that may impact the comparison.

 

[tornado99]

The person on Hacker News (or perhaps it was here) explicitly said that PDF rendering is independent of pixel-doubled vs non pixel-doubled. They must have some justification for that.

 

[motrek]

The person on Hacker News (or perhaps it was here) explicitly said that PDF rendering is independent of pixel-doubled vs non pixel-doubled. They must have some justification for that.

They may have meant that vector graphics are different from raster graphics and don't have an intrinsic resolution (until they're rasterized).

 

[wnyc1]

Nice thread.

 

[sautzer]

That's not true, otherwise it wouldn't be compatible with Apple Silicon laptops with only one external display stream available, such as the M1 MBA I'm using mine with.

Hi all, I've got a Macbook Air M1 512Gb/16Gb and I've got a PhilipsPerfectKolor 275P4VYKEB/00. Like your LG, it needs two displayport cables to achieve the 5k resolution, otherwise you'll only get 4k. In which way you connected your MBA to the monitor?

UPDATE 1: I've come across this two statements while browsing this forum:

• M1 Macs can run a dual-link SST display using a Thunderbolt 3 to Dual DisplayPort adapter despite only supporting one “normal” external display. There’s no need to bother with DisplayLink.
• M1 Macs can also run a dual-link SST display without a Thunderbolt adapter since they have more than one Thunderbolt port.

Is it advisable to buy always a TB3/TB4 dock? Otherwise, could I use a USB C hub with two USB C and double displayport?

Also it appears that I need .mtdd file to run the monitor but that is not available. Can someone help me make one? Also by linking documentation.

EDIT:

M1 is weird. While it can't support two separate displays from Thunderbolt, it does support a dual tiled display with two separate DisplayPort outputs from a Thunderbolt port.

The M1 Mac mini has 3 DisplayPort outputs. One for HDMI, one for Thunderbolt, and a third one for Thunderbolt that can only be used by a dual tiled display.

Thunderbolt 3/4 supports 5K60 10bpc RGB with dual tiles. You don't need deeper than that.
DSC can do 8K60 at 10bpc. It compresses 12bpc to 12bpp.

Sorry to tag also you but at this point, to drive my monitor, I need a thunderbolt 3/4 that has dual HDMI/DP ports? How can I choose the right one?
@joevt

 

[bllshttng]

Yeah the menu rotation option was supposed to be used to allow you to have the menu at the bottom of the screen regardless of the monitor’s physical orientation. Apparently it was buggy though so they removed it. One of my monitors came with the firmware with that option and I found that it woke significantly faster than the one that didn’t have it. Also, the USB drive will have a Firmware report that will tell you the firmware date if you are interested in that. As far as the USB hub, both of mine work as expected when connected via USB-C, they are both only USB 2.0 though.

Can you tell me more about the menu rotation option? What does it enable? My menu is really small and in the bottom right corner.

The actual monitor is relatively easy to rotate to portrait mode (like Pro Display XDR), but there is obviously no “sensing” of the orientation change that is passed along to the OS (so you just have to change it in display settings).

Is the USB hub working for you? I only tried with one (ancient) accessory which didn’t work.

Where are you finding firmware updates? I checked the clickclack website and see that there's translated info about the firmware but no links for downloads.

I just got 2 Kuycon 5K monitors and so far they are pretty great looking. I think I need to configure with BetterDisplay to see if I can try to match the icc profile from my macbook pro m2 max's display.

BTW, There's a few more monitors available on clickclack.io if anyone is in the market for a 5K monitor that's cheaper than ASD and ViewFinity S9 and has a glossy display.

Excuse the mess, but here's some details and closeups on the look and feel of the Kuycon's. I also have a caldigit ts4, but I'm not sure if it can drive the 2 5K monitors natively. So far, I'm realizing that I lose HDR ability in one of the monitors if I drive both through the TS4 (it returns if I plug one monitor in directly into the macbook and plug another via ts4).

 

[cinedog]

Where are you finding firmware updates? I checked the clickclack website and see that there's translated info about the firmware but no links for downloads.

I just got 2 Kuycon 5K monitors and so far they are pretty great looking. I think I need to configure with BetterDisplay to see if I can try to match the icc profile from my macbook pro m2 max's display.

BTW, There's a few more monitors available on clickclack.io if anyone is in the market for a 5K monitor that's cheaper than ASD and ViewFinity S9 and has a glossy display.

Excuse the mess, but here's some details and closeups on the look and feel of the Kuycon's. I also have a caldigit ts4, but I'm not sure if it can drive the 2 5K monitors natively. So far, I'm realizing that I lose HDR ability in one of the monitors if I drive both through the TS4 (it returns if I plug one monitor in directly into the macbook and plug another via ts4).

Nice pictures! I was wondering, could you post a close up of one of the black holes on the back? I'm curious if the holes are purely cosmetic, or if there are small perforations/holes inside each of the metal holes that help heat escape the monitor. I'm wondering if the hole version dissipates heat better than the solid metal back version.

 

[bllshttng]

Nice pictures! I was wondering, could you post a close up of one of the black holes on the back? I'm curious if the holes are purely cosmetic, or if there are small perforations/holes inside each of the metal holes that help heat escape the monitor. I'm wondering if the hole version dissipates heat better than the solid metal back version.

Here you go. I don't think I've used it enough to feel any heat escape, but it does have perforations within each of the cut out holes.

 

[yfyfyf]

so glad I found this thread, I just want to report on a 5k monitor that I found last month. They have stock and I am able buy the unit from T-mall here in Hong Kong. The monitor itself doesn't support VESA mounts but the word is that a new revision with that option will out soon.

YouTube:
2999的苹果Studio Display平替？未来视野RV100 5K显示器测评YouTubehttps://www.youtube.com › watch
product page:
RV100 - RichVision未来视野官方网站richvision.comhttp://www.richvision.com › product

 

[EugW]

so glad I found this thread, I just want to report on a 5k monitor that I found last month. They have stock and I am able buy the unit from T-mall here in Hong Kong. The monitor itself doesn't support VESA mounts but the word is that a new revision with that option will out soon.

YouTube:
2999的苹果Studio Display平替？未来视野RV100 5K显示器测评YouTubehttps://www.youtube.com › watch
product page:
RV100 - RichVision未来视野官方网站richvision.comhttp://www.richvision.com › product

Nice. Like many others though, this stuck with HDMI 2.0 instead of using HDMI 2.1. I guess the panel can’t support higher refresh rates anyway.

 

[motrek]

so glad I found this thread, I just want to report on a 5k monitor that I found last month. They have stock and I am able buy the unit from T-mall here in Hong Kong. The monitor itself doesn't support VESA mounts but the word is that a new revision with that option will out soon.

YouTube:
2999的苹果Studio Display平替？未来视野RV100 5K显示器测评YouTubehttps://www.youtube.com › watch
product page:
RV100 - RichVision未来视野官方网站richvision.comhttp://www.richvision.com › product

Looks great. How much did it cost?

 

[CWallace]

It is interesting we are seeing a sudden explosion of third-party 5K monitors. BOE is now making a 5K panel, as well, so I presume many of these are using that panel and it has better (lower) pricing than the LG panel Apple has been using since 2015.