Samsung Odyssey Neo G7 (S32BG75)

Interlace pattern artifacts

On some monitors, particularly but not exclusively those with high refresh rates, interlace patterns can be seen during certain transitions. We refer to these as ‘interlace pattern artifacts’ but some users refer to them as ‘inversion artifacts’ and others as ‘scan lines’. They may appear as an interference pattern, mesh or interlaced lines which break up a given shade into a darker and lighter version of what is intended. They often catch the eye due to their dynamic nature, on models where they manifest themselves in this way. Alternatively, static interlace patterns may be seen with some shades appearing as faint horizontal or vertical bands of a slightly lighter and slightly darker version of the intended shade.

We observed some static interlace patterns on the Odyssey Neo G7, but they weren’t extreme by any means. Some shades appeared as alternating horizontal bands of a slightly lighter and darker version of the intended shade. This was observed for various shade including certain blue, green, cyan and paler orange or yellow shades. This didn’t affect the entire screen equally and tended to be more noticeable near the bottom and also side edges (more so the right side). There seemed to be some interaction with viewing angle as these could be masked or appear more prominent in certain sections of the screen as viewing angle was adjusted. They were clearest at 165Hz but still reasonably faint here and something most won’t notice. They were even fainter at 120Hz and not observable as horizontal lines at 60Hz. The Neo G8 by comparison shows much clearer static interlace patterns when running at 240Hz.

The image below shows the medium blue shade which highlights selected resolutions in Nvidia Control Panel at 60Hz, 120Hz and 165Hz. The text is different at 60Hz compared to higher refresh rates because it’s pulled from a larger photo which has the resolution highlighted and the resolution is referenced differently by Nvidia. The target blue shade is identical in all cases. At 165Hz (bottom bar) you can see alternating horizontal stripes of lighter and darker shade, which are quite faint to the eye. At 60Hz (top bar) these stripes aren’t visible, but the shade shows a patchwork of static dithering where the screen alternates between a lighter and darker version of the intended shade within the same row of pixels. To the eye this was again faint but could be observed in place of the static interlace patterns that were seen at higher refresh rates. At 120Hz (middle bar) things appeared as a sort of blend of the two patterns and to the eye really just appeared as very faint horizontal lines.

The text clarity may appear different at various refresh rates in these images, but this is not observed by eye and is just due to differences in camera focus or processing. Several shots were taken at each refresh rate and the characteristic patterns discussed here for the blue background were observed in the way shown and described at a given refresh rate.

Static interlace patterns, various refresh rates

Dynamic ‘interlace pattern artifacts’ were also observed. To the eye it was as if the dithered pattern at 60Hz appeared to shift if you moved your eyes or observed moving content on the screen. It could be described as an interference pattern, grid or mesh. This could be observed on any section of the screen and also for shades which don’t clearly show static interlace patterns. It was also faintly visible at higher refresh rates, but much less readily observed. The patterns were observed with VRR enabled in much the same way, with their appearance depending on the refresh rate and whether that was closer to 60Hz (or below) or 165Hz for example. Although we like to mention these interlace pattern artifacts (both static and dynamic), that’s only for completeness and we’d stress that most people will not notice these. They aren’t particularly strong or obvious in this case, regardless of refresh rate.

Responsiveness

Input lag

A sensitive camera and a utility called SMTT 2.0 was used to analyse the latency of the Odyssey Neo G7. Over 30 repeat readings were taken to help maximise accuracy. Using this method, we calculated 3.45ms (slightly above ½ a frame at 165Hz) of input lag. Similar values were recorded at 120Hz and 60Hz – and with ‘Local Dimming’ enabled. These figures were recorded using the ‘Low Input Lag’ setting in the ‘Game’ section of the OSD. If you use VRR, which requires ‘Adaptive-Sync’ or ‘AMD FreeSync Premium Pro’ to be enabled in the OSD, the setting is not available. Under VRR we recorded a higher but still reasonable value of 4.31ms with ‘VRR Control’ off and a significantly higher 9.87ms with ‘VRR Control’ on. These figures are influenced by both the element of input lag you ‘see’ (pixel responsiveness) and the main element you ‘feel’ (signal delay). They indicate a low signal delay with VRR off and reasonably low signal delay with VRR on (‘VRR Control’ off) which most users should find acceptable. Note that we don’t have the means to accurately measure input lag with VRR technology active in a VRR environment or HDR active in an HDR environment.

Perceived blur (pursuit photography)

Our article on responsiveness explores the key factors related to the responsiveness of monitors. An important concept detailed in the article is ‘perceived blur’, which is contributed to by both the pixel responses of the monitor and the movement of your eyes as you track motion on the screen. This second factor is predominant on modern monitors, though both factors play an important role. A photography technique called ‘pursuit photography’ is also explored, which uses a moving rather than stationary camera to capture motion in a way that reflects both elements of perceived blur. Rather than solely reflecting the pixel response component.

The images below are pursuit photographs taken using the UFO Motion Test for ghosting, with the test running at its default speed of 960 pixels per second. This is a good practical speed to take such photographs at and highlights both elements of perceived blur well. The UFOs move across the screen from left to right at a frame rate matching the refresh rate of the display. All three rows of the test are analysed to highlight a range of pixel transitions. The monitor was tested at 60Hz (directly below), 120Hz and 165Hz. The three main ‘Response Time’ settings are tested as well as VRR active, which greys this out so it’s unselectable. The two final columns show reference screens, set to what we consider their optimal response time setting for a given refresh rate. The AOC CQ32G3SU, which is another 165Hz VA model that provides quite an average pixel response performance for the panel type. And the Gigabyte M27Q, which is an IPS model offering a decent but not outstanding pixel response experience – one which most are perfectly happy with.

At 60Hz, above, the UFO appears soft and unfocused without clear internal detailing. This reflects a moderate amount of perceived blur due to eye movement. Some trailing is observed due to some weaknesses in pixel response times or overly aggressive tuning. ‘VRR’ enabled and the ‘Standard’ response time setting offer a similar performance and show less clear weaknesses (less obvious ‘powdery’ trailing) compared to the CQ32G3SU, particularly for the dark background (top row) and medium background (middle row). Performance for the light background (bottom row) is impressive and also better than the CQ32G3SU. There is a bolder initial trail than with the M27Q for the dark and a lesser extent medium background due to some relative pixel response time weaknesses for these particular transitions. The ‘Faster’ and moreover ‘Extreme’ settings introduce clear overshoot (inverse ghosting), including colourful and bright ‘halo trailing’ which is brighter than the background and stands out for that reason. And some inky or ‘dirty’ trailing that’s darker than the background, most noticeable for behind the yellow UFO cockpit area for the dark background. We consider ‘Standard’ (or VRR) optimal here. Below you can see what things look like with refresh rate doubled to 120Hz.

At 120Hz, above, the UFO appears significantly narrower with clearer internal detail. This reflects a significant decrease in perceived blur due to eye movement. To the eye the black lines which separate the segments of the UFO are less blended than they appear in the photos, so segmentation is more distinct. The pixel response requirements for a good performance here are significantly increased. The monitor ramps up the aggressiveness of its pixel overdrive. You can see some overshoot even using the ‘Standard’ setting or VRR, mainly as slight ‘dirty’ trailing which gives a bit of a shadow behind the UFO cockpit for the dark and medium background. There is also some conventional trailing, with bold initial trail for the dark background in particular, behind the red UFO body. We might describe what is seen here for the dark background as ‘heavy powdery’ trailing. But it doesn’t have the smeary appearance of the trailing for the CQ32G3SU reference (which plagues most VA models) and is not as much of a significant or noticeable weakness in practice. The M27Q doesn’t have the same bold initial component to the trail or overshoot, but has ‘powdery’ trailing which extends further back – a fairly typical pixel response time weakness for IPS models. The ‘Faster’ and ‘Extreme’ settings both show more noticeable overshoot, with a bright and colourful appearance in places and a deeply inky look elsewhere. We again consider ‘Standard’ (or VRR) optimal here. Below you can see things bumped up a bit further, to 165Hz.

At 165Hz, above, the UFO appears slightly narrower with somewhat improved definition. To the eye the segmentation is quite distinct now, though the white notches still appear somewhat blended as they appear in the photos. This reflects a further reduction in perceived blur to eye movement. The pixel response requirements for a good performance are further bumped up here. Things shift a bit more towards conventional trailing in place of overshoot compared to at 120Hz. The ‘Standard’ setting or VRR shows a fairly bold initial trail, extending just slightly for the dark background. And there’s a touch of overshoot in places. But the weaknesses are far less pronounced than with the CQ32G3SU – which is itself far from the worst ‘high refresh rate’ VA performer we’ve come across. The M27Q has more distinct ‘powdery’ trailing behind the UFO cockpit in particular, but has lower overshoot levels and has a less bold initial component to the trail behind the UFO body for the dark and medium backgrounds. The ‘Faster’ and ‘Extreme’ settings again show more noticeable overshoot and for some transitions not shown we found this very eye-catching. We again consider ‘Standard’ (or VRR) optimal here. Though not perfect, the Samsung performs impressively well for a VA panel here.

The monitor includes a strobe backlight or ‘Motion Blur Reduction’ setting we introduced earlier with respect to brightness; ‘Extreme (MBR)’. This is listed as a ‘Response Time’ setting with VRR and ‘Local Dimming’ disabled. This setting forces the backlight to flicker in sync with the refresh rate of the display, with 60Hz, 120Hz or 165Hz selectable. 60Hz being selectable could be unintentional as the setting doesn’t work as it should – the monitor strobes at 120Hz and this gives very poor motion clarity with the monitor set to 60Hz. The flickering of this strobe backlight setting can accelerate eye fatigue even if the flickering isn’t actively noticed, but even with this in mind some users still like the competitive edge such a setting can bring in terms of minimising perceived blur. Unfortunately, there are significant strobe timing issues with this setting. The level of strobe crosstalk (duplication of the main object) is significant at 165Hz. It’s about as bold as the actual object, it overlaps it and makes it appear as if the object is duplicated. If you’re gaming, this would apply to various objects, textures or other asset in the game – they would appear to duplicate during movement due to the heavy strobe crosstalk. At 120Hz the strobe crosstalk constantly shifts between being behind and in front of the object, crossing over the main object as it does so – which is extremely weird and distracting. We consider this feature essentially unusable or at least very undesirable because of this and it fails to achieve its main objective – so we won’t be including our full assessment of the setting. Below you can see how things look using ‘Test UFO’ and this setting at 120Hz and 165Hz. The strobe crosstalk is shown as displaced in front and behind at 120Hz, as described previously. Remember that the UFOs move across the central bands of the screen during this test, where your eyes mainly focus during competitive gameplay and where you’d ideally see low strobe crosstalk.

Responsiveness in games and movies

On various Battlefield titles, at a frame rate keeping up with the 165Hz refresh rate, the monitor delivered a fluid experience. Compared to a 60Hz monitor or this monitor running at 60Hz (or 60fps), the monitor provides 2.75 times as much visual information every second. This has two main effects. It greatly improves the ‘connected feel’, which describes the precision and fluidity felt when interacting with your character or the game environment. This is also aided by the low input lag of the screen. The high frame and refresh rate combination also decreases perceived blur due to eye movement, something we demonstrated earlier using Test UFO. These advantages can not only make the gaming experience more comfortable for some, it can also provide a competitive edge in games like Battlefield. Making it easier to track and engage enemies. It complements the high pixel density of the screen well, too, with better preservation of some of the detail during motion due to the increased frame and refresh rate.

Pixel responsiveness is another key part of perceived blur and that is the traditional Achilles heel of VA technology. As demonstrated earlier using Test UFO, the monitor is not perfect in that respect but is also very impressive for the panel type. Most transitions are performed very quickly – those exclusively involving medium to brighter shades exceptionally so, without particular weaknesses of note due to slower than optimal pixel responses. Where darker shades are introduced, some more significant weaknesses are introduced – though these remain very much on the minor side for a VA model. There is ‘powdery’ trailing which we might describe as ‘heavy’ or ‘bold’ for the slowest transitions (certain dark shades against medium backgrounds). But this doesn’t take on the distinct ‘smeary’ appearance typical of VA models. Whilst some fast IPS models wouldn’t exhibit the same weaknesses for these transitions, others would – and some are actually worse in this respect. So really, this is a competent performance which is particularly impressive given the panel type used. It was also free the ‘flickering effect’ some VA models have with mixtures of lighter and darker shades where the brighter shade dims during movement and brightens when movement ceases.

There was reasonably low overshoot overall – a bit of ‘dirty’ trailing and ‘halo’ trailing as we described with Test UFO but nothing that we found really jumped out in an eye-catching way for most transitions. There was some slightly eye-catching overshoot where some dark shades were involved, but this was still far from extreme. We made similar observations on Shadow of the Tomb Raider, so a fluid 165Hz experience overall. There are plenty of darker shades mixed in on this title, highlighting some of the weaker transitions performed. But again, the typical VA ‘smears’ were absent here and there wasn’t strong overshoot in its place, either. We also observed video content at a range of refresh rates, including ~24 – 30fps content on platforms such as Netflix and 60fps content on YouTube. There were no major weaknesses from the pixel responses of the monitor. There was just a touch of ‘powdery’ (nowhere close to ‘smeary’) trailing for some transitions involving certain dark shades for the 60fps content. And some traces of overshoot in places, but nothing that really stood out in a clear way. The main barrier to fluidity was the frame rate of the content itself rather than weaknesses related to the pixel responses of the monitor.

As an Amazon Associate I earn from qualifying purchases made using the below link. Where possible, you’ll be redirected to your nearest store. Further information on supporting our work.

VRR (Variable Refresh Rate) technology

FreeSync – the technology and activating it

AMD FreeSync is a variable refresh rate technology, an AMD-specific alternative to Nvidia G-SYNC. Where possible, the monitor dynamically adjusts its refresh rate so that it matches the frame rate being outputted by the GPU. Both our responsiveness article and the G-SYNC article linked to explore the importance of these two elements being synchronised. At a basic level, a mismatch between the frame rate and refresh rate can cause stuttering (VSync on) or tearing and juddering (VSync off). FreeSync also boasts reduced latency compared to running with VSync enabled, in the variable frame rate environment in which it operates.

FreeSync requires a compatible AMD GPU such as the Radeon RX 580 used in our test system. The monitor itself must support ‘VESA Adaptive-Sync’ for at least one of its display connectors, as this is the protocol that FreeSync uses. The Odyssey Neo G7 supports FreeSync Premium Pro via DP and HDMI on compatible GPUs and systems. Note that HDR can be activated at the same time as FreeSync. You need to make sure ‘FreeSync Premium Pro’ is enabled in the ‘Game’ section of the OSD. On the GPU driver side recent AMD drivers make activation of the technology very simple. You should ensure the GPU driver is setup correctly to use FreeSync, so open ‘AMD Software’, click ‘Settings’ (cog icon towards top right) and click on ‘Display’. You should then ensure that the first slider is set to ‘Enabled’ as shown below. The top image shows the monitor connected by DP and the bottom image by HDMI. The setting is referred to as ‘AMD FreeSync Premium Pro’ in both cases, although the exact wording may depend on the driver version you’re using.

The Samsung supports a variable refresh rate range of 48 – 165Hz. That means that if the game is running between 48fps and 165fps, the monitor will adjust its refresh rate to match. When the frame rate rises above 165fps, the monitor will stay at 165Hz and the GPU will respect your selection of ‘VSync on’ or ‘VSync off’ in the graphics driver. With ‘VSync on’ the frame rate will not be allowed to rise above 165fps, at which point VSync activates and imposes the usual associated latency penalty. With ‘VSync off’ the frame rate is free to climb as high as the GPU will output (potentially >165fps). AMD LFC (Low Framerate Compensation) is also supported by this model, which means that the refresh rate will stick to multiples of the frame rate where it falls below the 48Hz (48fps) floor of operation for FreeSync. If a game ran at 35fps, for example, the refresh rate would be 70Hz to help keep tearing and stuttering at bay. LFC sometimes seemed to activate at slightly higher refresh rates, sometimes as high as 53Hz (53fps) – this slightly different floor of operation makes little difference in practice. This feature is used regardless of VSync setting, so it’s only above the ceiling of operation where the VSync setting makes a difference.

To configure VSync, open ‘AMD Software’. Click ‘Settings’ (cog icon towards top right) and click ‘Graphics’. The setting is listed as ‘Wait for Vertical Refresh’. This configures it globally, but if you wish to configure it for individual games click ‘Game Graphics’ towards the top right. The default is ‘Off, unless application specifies’ which means that VSync will only be active if you enable it within the game itself, if there is such an option. Such an option does usually exist – it may be called ‘sync every frame’ or something along those lines rather than simply ‘VSync’. Most users will probably wish to enable VSync when using FreeSync to ensure that they don’t get any tearing. You’d therefore select either the third or fourth option in the list, shown in the image below. Above this dropdown list there’s a toggle for ‘Radeon Enhanced Sync’. This is an alternative to VSync which allows the frame rate to rise above the refresh rate (no VSync latency penalty) whilst potentially keeping the experience free from tearing or juddering. This requires that the frame rate comfortably exceeds the refresh rate, not just peaks slightly above it. We won’t be going into this in detail as it’s a GPU feature rather than a monitor feature.



Some users prefer to leave VSync enabled but use a frame rate limiter set a few frames below the maximum supported (e.g. 162fps) instead, avoiding any VSync latency penalty at frame rates near the ceiling of operation or tearing from frame rates rising above the refresh rate. If you go to ‘Support’ – ‘Information’ in the OSD, the current refresh rate of the monitor is listed. This updates in real time and will reflect the frame rate of the content, if the monitor is within the main VRR window. Finally, it’s worth noting that FreeSync only removes stuttering or juddering related to mismatches between frame rate and refresh rate. It can’t compensate for other interruptions to smooth game play, for example network latency or insufficient system memory. Some game engines will also show stuttering (or ‘hitching’) for various other reasons which won’t be eliminated by the technology.

FreeSync – the experience

As always, various game titles were tested using AMD FreeSync and the experience was similar in all cases. Any issues which affect one title but not others would suggest a GPU driver issue or game issue rather than a monitor issue. To keep things simple, we’ll just focus on Battlefield titles for this section. The full VRR range of the monitor can be tested with these titles using our Radeon RX 580. This isn’t a very powerful GPU by modern standards, so it wasn’t unusual to see dips below 165fps. Without VRR technology in use, these dips would be accompanied by tearing (VSync off) or stuttering (VSync on) from frame and refresh rate mismatches. FreeSync removed these mismatches and if you’re sensitive to them, it’s certainly nice to have them removed. There are still downsides to reduced frame rate in the form of a worse ‘connected feel’ and increased perceived blur, however.

The monitor doesn’t allow the ‘Response Time’ setting to be adjusted with VRR active. The monitor employs a degree of variable overdrive, though, which reduces the aggressiveness of the pixel overdrive as refresh rate decreases – something which coincides with frame rate decreases when VRR is active. This ensures you don’t get a significant increase in overshoot over a broad range of transitions, at reduced refresh rates. But we did notice some ‘halo’ trailing around some darker objects with certain medium-bright shades in the background, which became more noticeable at decreased refresh rates. With drops to ~120Hz this was quite noticeable and it became increasingly noticeable as things dipped further into the double digits. By ~60Hz it was rather eye-catching and difficult to ignore – potentially quite annoying where affected transitions are present, which isn’t uncommon in darker scenes or potentially for shadowy areas of brighter scenes. Although the pixel overdrive was slackened off somewhat as refresh rate decreased, it still remained strong enough to provide a respectable performance across the VRR range. FreeSync worked down to the floor of operation (~48Hz – 53Hz), below which LFC (Low Framerate Compensation) kicks in. This keeps the refresh rate at a multiple of the frame rate to keep tearing and stuttering at bay. There was a subtle momentary stuttering when LFC activated or deactivated, something we always observe and not specific to this model. It was much less noticeable than traditional stuttering from frame and refresh rate mismatches, but if you’re sensitive to it and frequently passing the boundary it could be annoying.

An issue which requires some discussion with VRR active on this model is ‘VRR flickering’. Monitors with particularly strong contrast (such as this one) are prone to flickering for some shades during heavy frame rate fluctuations. These are due to slight gamma changes at the lower end of the curve for certain refresh rate changes in a VRR environment and are not visible on models with relatively weak contrast. VA models like this are also particularly sensitive to the voltage changes that occur during heavy refresh rate fluctuation, which can cause flickering not just for darker shades but elsewhere as well. We observed some flickering with VRR active, mainly during heavy fluctuations in refresh rate. If the frame rate stayed in the triple digits or dipped just below that, there wasn’t noticeable flickering (or if there was, it was on the mild side). It was a bit easier to notice but still not extreme in our view when the brightness was increased – including with ‘Local Dimming’ active under SDR or HDR. For significant fluctuations where double digit frame rates were involved, we noticed some flickering but it was usually fairly mild unless the fluctuation was rather extreme. It was particularly strong as the LFC boundary was crossed in either direction, as this involves a significant sudden increase or drop in refresh rate with double digit frame rates involved. Loading screens in some games and some in-game maps or menus could trigger moderate flickering. This monitor wasn’t the worst offender in terms of VRR flickering overall – some models show obvious flickering even when the refresh rate is relatively stable. But it’s certainly still something to consider if you’re sensitive to flickering and like to use VRR technologies like FreeSync.

The ‘VRR Control’ setting found in the ‘System’ section of the OSD gets rid of this flickering, but as measured earlier significantly increases input lag. It also introduces stuttering. Whilst this isn’t as obvious as the stuttering you can get with VRR disabled (VSync on) and there isn’t any tearing, we still found this stuttering rather obnoxious. It occurred even if the frame rate was high and also when it was very stable. If you’re particularly sensitive to flickering but less so stuttering and don’t mind the increase in input lag then this setting may be of use. But for us it detracts from the purpose of VRR technology with the stuttering that occurs and we prefer to keep ‘VRR Control’ disabled.

Nvidia Adaptive-Sync (‘G-SYNC Compatible’)

As noted earlier, AMD FreeSync makes use of Adaptive-Sync technology on a compatible monitor. As of driver version 417.71, users with Nvidia GPUs (GTX 10 series and newer) and Windows 10 or later can also make use of this Variable Refresh Rate (VRR) technology. When a monitor is used in this way, it is something which Nvidia refers to as ‘G-SYNC Compatible’. Some models are validated as G-SYNC compatible, which means they have been specifically tested by Nvidia and pass certain quality checks. With the Odyssey Neo G7 you can connect the monitor up via either DisplayPort 1.4 or HDMI 2.1 to use ‘G-SYNC Compatible Mode’, with the latter technically making use of HDMI 2.1’s integrated VRR functionality rather than Adaptive-Sync. You need to make sure ‘Adaptive-Sync’ is switched on in the ‘Game’ section of the OSD to use the technology via either DP or HDMI as this acts as a VRR toggle on this monitor. When you open up Nvidia Control Panel, you should then see ‘Set up G-SYNC’ listed in the ‘Display’ section. Ensure the ‘Enable G-SYNC, G-SYNC Compatible’ checkbox and ‘Enable settings for the selected display model’ is checked as shown below and press ‘OK’. If you’ve enabled ‘G-SYNC Compatible’ and it was previously disabled, the monitor should re-establish its connection with the system and the technology should now be active.

The image below is taken from another monitor – although the model number differs, the rest of the information in the image is identical.



You will also see in the image above that it states: “Selected Display is not validated as G-SYNC Compatible.” This means Nvidia hasn’t specifically tested and validated the display, not that it won’t work. On our RTX 3090 the experience was very similar to what we described with FreeSync. With the technology getting rid of tearing and stuttering from what would otherwise be frame and refresh rate mismatches, within the VRR range. Similar VRR flickering behaviour was observed, mainly during heavy fluctuations in frame rate and hence refresh rate. There appeared to be a slightly higher floor of operation of 55Hz, so 55 – 165Hz. Though an LFC-like frame to refresh multiplication technology was employed below that to keep tearing and stuttering from frame and refresh rate mismatches at bay. There was again a subtle momentary stuttering and quite noticeable flickering as the boundary was crossed, as we observed with our AMD GPU as well. Our suggestions regarding use of VSync also apply, but you’re using Nvidia Control Panel rather than AMD Software to control this. The setting is found in ‘Manage 3D settings’ under ‘Vertical sync’, where the final option (‘Fast’) is equivalent to AMD’s ‘Enhanced Sync’ setting. You’ll also notice ‘G-SYNC Compatible’ listed under ‘Monitor Technology’ in this section, as shown below. Make sure this is selected (it should be if you’ve set everything up correctly in ‘Set up G-SYNC’).



If you go to ‘Support’ – ‘Information’ in the OSD, the current refresh rate of the monitor is listed. This updates in real time and will reflect the frame rate of the content, if the monitor is within the main VRR window. And as with AMD FreeSync, HDR can be used at the same time as ‘G-SYNC Compatible’.

HDMI 2.1 VRR

HDMI 2.1 includes Variable Refresh Rate (VRR) support as part of the specification. This is an integrated technology, which unlike FreeSync does not rely on VESA Adaptive-Sync to function. As such it be used by devices such as the PS5 that don’t support Adaptive-Sync. It can also be leveraged via ‘G-SYNC Compatible Mode’ on compatible Nvidia GPUs. Although Adaptive-Sync isn’t used, you need to have ‘Adaptive-Sync’ active in the OSD of the monitor as this acts as a VRR toggle on this model. Based on our testing of ‘G-SYNC Compatible Mode’ using HDMI 2.1 VRR, the experience was very similar to the Adaptive-Sync experience under SDR and HDR.

HDR (High Dynamic Range)

HDR (High Dynamic Range) on an ideal monitor involves very deep dark shades and brilliantly bright light shades being displayed simultaneously. Alongside a broad spectrum of shades between these extremes, including muted pastel shades and highly vibrant saturated shades. Ideally, per-pixel illumination would be offered (e.g. self-emissive displays such as OLED), or for LCDs a very large number of precisely controlled dimming zones used. A solution such as FALD (Full Array Local Dimming) with a generous number of dimming zones, for example. Such a solution would allow some areas of the screen to remain very dim whilst others show brilliantly high brightness. Colour reproduction is also an important part of HDR, with the ultimate goal being support for a huge colour gamut, Rec. 2020. A more achievable near-term goal is support for at least 90% DCI-P3 (Digital Cinema Initiatives standard colour space) coverage. Finally, HDR makes use of at least 10-bit precision per colour channel, so its desirable that the monitor supports at least 10-bits per subpixel. HDR10 is the most widely supported standard used in HDR games and movies and what is supported here.

For games and other full screen applications that support HDR, the Odyssey Neo G7 automatically switches to its HDR operating mode if an HDR signal is provided. Some game titles will activate HDR correctly when the appropriate in-game setting is selected. Others that support HDR will only run in HDR if ‘Use HDR’ is turned on in Windows, too. Related Windows HDR settings are found in the ‘Windows HD Color settings’ (Windows 10) or ‘HDR’ (Windows 11) section of ‘Display settings’ (right click the desktop). If you want to view HDR movie content, ensure ‘Stream HDR Video’ (Windows 10) or ‘Play streaming HDR video’ (Windows 11) is active. Also note that there’s an ‘HDR/SDR brightness balance’ (Windows 10) or ‘SDR content brightness’ (Windows 11) slider that allows you to adjust the overall balance of SDR content if HDR is active in Windows. This is really just a digital brightness slider that only makes changes for SDR content, and you lose contrast by adjusting it. Image balance is upset when viewing SDR content under HDR, as if gamma is quite a bit too low overall which gives a rather foggy look and reveals a fair bit of unintended detail – colour accuracy also suffers. As usual. we’d recommend only activating HDR in Windows if you’re about to use an HDR application that specifically requires it.

For simplicity we’ll just focus on Shadow of the Tomb Raider for this section. This is a title we’ve tested extensively, observed on a broad range of monitors under HDR. It has a good HDR implementation which is very much limited by the screen itself, highlighting strengths and weaknesses in HDR performance well. Although our testing focuses on HDR PC gaming using DisplayPort on an RTX 3090, similar observations were made when viewing HDR video content on the Netflix app. As with games, some HDR video content makes better use of HDR than others. There are some additional points to bear in mind if you wish to view such content. We also made similar observations using HDMI, which would be used when viewing HDR content on an HDR compatible games console for example. Testing on both our Nvidia and AMD GPUs showed that the HDR implementation was similar in both cases, too.

Unusually for HDR, many of the settings remain available for you to change including brightness, gamma and colour channels. Any adjustments you make in the ‘Picture’ section of the OSD apply to HDR but won’t carry over if you switch to SDR. We wouldn’t recommend making significant adjustments here as it will go against the intended HDR output. It’s fine to make slight colour channel adjustments as you might under SDR to rebalance white point or get rid of an obvious tint. Reducing brightness a bit if you’re sensitive to it and prefer things to look dimmer is also ‘OK’, but significant adjustments there really make things look rather dull and faded. The monitor is really optimised and configured under HDR for high brightness levels. The main setting of interest under HDR is ‘Local Dimming’, found in the ‘Game’ Section of the OSD. As highlighted earlier with respect to brightness, there are three settings here; ‘Off’, ‘Auto’, ‘Low’ and ‘High’. ‘Auto’ and ‘High’ are quite similar to one another overall, or certainly were on our unit and in our testing. The ‘High’ setting seems to slightly brighten some small bright highlights, helping them pop out just a bit more in some instances. Overall, we didn’t find the difference between the two settings dramatic. The ‘Low’ setting limits brightness levels significantly and favours dimming over brightening zones where possible. It also doesn’t dim as aggressively for the darkest shades. We certainly felt the ‘Low’ setting limited the potential of the monitor too much under HDR, so we settled for ‘High’ for our testing here. Aside from that, we left things at default for our testing under HDR.

The Samsung Odyssey Neo G7 doesn’t include any specific VESA DisplayHDR certification. Samsung went off-piste with their own unique ‘Quantum HDR 2000’ marketing. This relates to the use of a 1196-zone Quantum Dot Mini LED backlight solution which is supposed to achieve a peak luminance of 2000 cd/m². Whilst the peak luminance levels we recorded earlier were very respectable, they still fell well short of that figure. We took some in-game measurements in Shadow of the Tomb Raider and a few other titles and regardless of how HDR was calibrated in the game, we did not record anywhere near this brightness level and didn’t exceed the sort of figures provided earlier. It may be possible to achieve this under extremely specific test conditions, but we’re more interested in what the user will see when using the monitor than misleading marketing. The use of QD LEDs for the backlight is how the monitor achieved its wide gamut, with 91% DCI-P3 coverage recorded. This is shown in the representation below, where the red triangle shows the monitor’s colour gamut, the blue triangle DCI-P3 and green triangle sRGB.

Colour gamut 'Test Settings'