Samsung Odyssey Neo G8 (S32BG85) review notes

Viewing 13 posts - 1 through 13 (of 13 total)

Buying a monitor? Please refer to this post before purchasing.
New user? Register here.


  • Author
    Posts
  • #68210
    PCM2

      This video review of the Odyssey Neo G8 (S32BG85) has been published by ‘Little Snowman’.

      It’s in Chinese but some interesting pointers from the video:

      – Monitor based on CELL (panel without backlight arrangement) from CSOT, the CY-PB315GLHV1H.

      – Stand was criticised for being rather shaky and unstable, not providing a solid feel to the monitor.

      – Uniformity of the review sample was good overall, though bright uniformity was affected by some ‘pinching’ at the edges. Sometimes referred to as ‘vignetting’ and certainly not unique to this model.

      – Normal sRGB subpixels without partial (‘split’) illumination, but ‘static interlace patterns’ (‘scan lines’ or what he called a ‘grid effect’) observed >120Hz (definitely 240Hz). Not sure about ‘dynamic interlace patterns’, they weren’t explicitly mentioned but could still be there. Hardware Unboxed (HUB) reviewed the Neo G7 (S32BG75) and didn’t notice this issue at its maximum 165Hz refresh rate.

      – DP 1.4 with DSC supports 3840 x 2160 @240Hz with 10-bit colour. HDMI 2.1 provides 40Gbps bandwidth (DSC also supported), providing up to 12-bit (scaler dithering) at native resolution and 240Hz. HDMI 2.1 VRR included, confirmed to work with PS5 with 48 – 120Hz VRR range. Provides a ‘4K’ 120Hz signal for PS5 and Xbox Series X.

      sRGB emulation mode included with ~99% coverage measured with little over-coverage. Native gamut is ~92% DCI-P3. Be careful when trying to interpret gamut measurements, some given in the video are ‘relative’ rather than ‘absolute’ values which don’t account for under-coverage in the usual way. He measured 130.9% sRGB, so Adobe RGB coverage would be fairly limited (~90%?) as usual for a Samsung QD LED model.

      4024:1 static contrast ratio recorded. ~15 cd/m² minimum luminance and ~355 cd/m² maximum luminance under SDR. Pretty decent adjustment range there with an impressively low minimum.

      – With local dimming active ‘PWM’ is used below grey levels of 170 (so medium-light grey and darker) with a ~957Hz cycling frequency. This doesn’t involve a complete turning off of the backlight but rather a dip in brightness, so is what we might refer to as ‘high frequency low amplitude oscillation’ rather than traditional PWM. Shouldn’t be bothersome to most people, could cause issues for those sensitive enough to flickering. Should broadly be considered a ‘flicker-free’ monitor as advertised.

      Local dimming with its 1196 zones appears to work well overall – 46 x 26 zones (H x V). Quite responsive and able to give a dynamic look to mixed content under HDR in particular, keeping strong depth where darker shades dominate and good bright ‘pop’ to brighter elements displayed at the same time. The local dimming test shown ~4 mins in was interesting. Seemed to suggest the algorithm is set up to bright-bias to keep brighter shades nice and bright at the expense of darker shades. Can be adjusted a bit in the OSD if you prefer to dark-bias things, but has a broader effect and can dim somewhat larger bright areas too much. Having an algorithm that dark-biases only for particularly small areas of bright shade surrounded by darkness and bright-biases otherwise might be preferred. PG32UQX tended to be too aggressive with its brightening even for very small bright areas, too. Some ‘haloing’ as expected, doesn’t appear extreme but stronger off-angle and perhaps when viewing peripheral regions of the screen.

      – Claimed 2000 cd/m² for HDR achieved and exceeded (2663 cd/m²) for a 10% window. ~953 cd/m² for 1% window and ~1408-1464 cd/m² for 3-4% window recorded. Dropped to ~683 cd/m² for 50% window, ~464 cd/m² for 75% window and ~450 cd/m² for 100% window. Compare this to PG32UQX which is able to sustain ~1200 cd/m²+ for a 100% window. Basically this shows a fairly aggressive Automatic Brightness Limiter (ABL) on the Neo G8 and will mean there’s noticeable dimming in scenes where bright shades dominate. Think bright daylight scenes with plenty of sky, sun-kissed landscapes etc.

      – HDR performance broadly considered ‘much better’ than Odyssey Neo G9 by reviewer.

      – Pixel responsiveness broadly echoes the 240Hz Odyssey G7 models. So very impressive for VA, though still some weaknesses (~10ms) for transitions involving dark background shades. To use my preferred terminology, that won’t produce ‘smeary’ trailing, but moderate ‘powdery’ trailing. Since I compared to the PG32UQX earlier, there’s really no contest there – the Neo G8 has faster pixel responses.

      – ‘Response Time’ setting (pixel overdrive adjustment) locked under VRR (Adaptive-Sync enabled). This can invite stronger overshoot at lower refresh rates – so reduced frame rates in a VRR environment – which can’t be counteracted.

      #69028
      PCM2

        Hardware Unboxed (HUB) has now posted a review of the S32BG75 – the Neo G7 or 165Hz version of the Neo G8. This review basically confirms that a lot of the above applies to the 165Hz version as well. They also confirmed poor ‘out of the box’ calibration with respect to gamma (and colour temperature), whilst they measured a slightly higher 93.5% DCI-P3 gamut with their setup. Reviewer Tim didn’t really get on with the 1000R curve and found it detrimental for productivity in particular and didn’t feel it added anything. He felt it was misplaced on a ~32″ 16:9 screen, something I can relate to and a viewpoint he’s aired on previous videos. He didn’t observe issues with ‘scan lines’ or VRR flickering during his testing even without using ‘VRR Control’. Responsiveness is strong overall, though is more tuned with 165Hz and similar high refresh rates in mind. There isn’t ‘smeary’ trailing to worry about, a bit of ‘powdery’ trailing as you can see from measurements and in the pursuit photos HUB includes – but overall strong pixel responsiveness. Effective variable overdrive isn’t included and there is no overdrive control in the OSD – so you will get some inescapable overshoot at low refresh rates.

        He much preferred the HDR experience to the Neo G9 and really liked its HDR performance overall. There weren’t weird quirks or unexpected issues as he observed on the Neo G9 and only really considered ‘blooming’ a particular issue of note for small bright elements immediately surrounded by very dark shades. He found the algorithm sometimes biased more towards dark than bright – so smaller bright elements are sometimes dragged down a bit rather than their darker background having raised depth. He noted HUDs in games sometimes appeared noticeably dimmer than they should because they were dragged down. Other times, perhaps for somewhat larger bright areas, he found the bright shades nice and bright with plenty of ‘pop’. He measured only 345 cd/m² for a 100% window (sustained), 431 cd/m² for a 75% window and 620 cd/m² for a 50% window, which echoes the subjective thoughts with the Neo G8 above regarding things dimming noticeably where bright shades dominate. He reported 1230 cd/m² for a ~10% window which is certainly bright but not up to the manufacturer’s brightness claims. He managed to record ~2300 cd/m² only for what he called a ‘pure 10% window’, but basically if this was modified even slightly it would drop down a lot. This is something reviewers of Samsung TVs have noted before, where they can show peak performance for specific test patterns such as a pure 10% window but in actual content will drop down a lot below that. Likely as they aren’t really designed to run at such a high brightness level very often if at all in actual normal content.

        I should also add that he didn’t come across any particular ‘Quality Control’ (QC) issues of note and the sample he looked at was just a random retail sample rather than specially selected by Samsung. Though I’ve personally received plenty of review samples that have been roughly handled by others and are far from cherry-picked despite what people might want to believe. Either way, this contrasted with his findings and general feedback received on the earlier Odyssey models – but this is only one sample and short-term usage, so I wouldn’t read too much into that. As an aside, he also picked up on the ‘poor viewing angle performance’. Though he didn’t specifically speak of colour consistency, or viewing angle weaknesses from a normal viewing position, the weak viewing angle performance off-angle usually manifests with colour consistency and gamma consistency weaknesses from a normal viewing position. These weaknesses are entirely usual for VA technology and this model does not seem ‘special’ or worthy of any particular praise in that respect. I personally feel it’s important to bear this in mind when comparing this model to something like the ASUS PG32UQX or even Sony INZONE M9. That, plus the significantly wider colour gamut (more suitable for HDR in that respect), will mean that the vibrancy or ‘pop to colours’ simply isn’t as strong on the Samsung when compared to the ASUS. This isn’t something reviewers will often convey as it requires a fair amount of subjective testing and analysis, but it’s definitely something to be aware of.

        #69041
        PCM2

          HUB has now completed the series with a review of the Neo G8 model. Embedded below for your convenience, but I will just add a few points I picked up on from the video. A lot of it echoes the above. In this case reviewer Tim did readily notice ‘scan lines’ (static interlace pattern artifacts) when observing certain shades at the native 240Hz, mainly for blue-biased shades. He also noted that the monitor supports 120Hz and 240Hz but nothing between the two. It was again a random retail sample purchased from Amazon and as you can see in the video (~7 mins) his sample had a yellowish brown splotch or “burn mark” as he called it at the bottom which would trigger an RMA of the unit. His G8 unit had a wonky gamma curve, unlike his G7 which was pretty good in that respect. And though the monitor appears impressively fast for a VA model ‘by the numbers’ (ignoring measured overshoot), I find the pursuit photographs (~14 mins) quite revealing. Depending on sensitivity to such things, the overshoot could be a bit bothersome and as with the G7 there is no overdrive control to change the tuning of that. And there are still some weaknesses from slower than optimal pixel responses mixed in, so it’s not a perfect 240Hz performance by any means.

          #69053
          jasswolf

            My impression of the HUB review is that the defective model may have been testing poorly for returning to near black, resulting in the exceptionally poor motion pursuit results for a fast 240Hz panel.

            These results on Reddit paint a picture that’s more competitive with the original G7.

            #69066
            PCM2

              Yeah, that’s more what I’d expect to see. Good spot. I have also found HUB’s pursuit photographs sometimes show more distinct trailing for some models I’ve also tested. And I know the pursuit photos I put in reviews very closely reflect what the eye sees. So perhaps the camera settings used or other conditions could be slightly odd as well and capture some things the eye wouldn’t. There could be a difference in warmup period in this case, too – I think if there’s genuinely this level of inter-unit variation with the Neo G8 then that isn’t a good thing.

              #69145
              PCM2

                Just another quick point not raised above – the Neo G7 has a lighter matte surface than the Neo G8. So glare handling isn’t quite as effective, but light emission from the monitor is more direct (less layering in front of the image, a ‘glassy’ look in some lighting etc.)

                #69389
                Gigamike

                  There are too many variables to say. Only a real comprehensive look at the variable refresh can say. It does look like there are differences, but only slightly. Could be panel variation but its very clear that you do get some dark smearing on both panels. Even the reddit post shows it with VRR enabled. At 120 hz its even more noticeable which really makes me worry what it would look like at even lower refresh rates.

                  The reddit pic (Neo G8 Motion)

                  Comparing it to the dell alienware AW3423DW @120 hz pic from PCM review

                  You can see clearly see the black blur behind the UFO.

                  We also dont know the settings used to do the motion blur test. Theres also camera variation and actual motion speed used to test both. Faster motion will usually result in longer trails, so its something that has to be taken into account here.

                  Of course it could be panel variation, but that worries me aswell because thats clearly too much variation for my taste. Though the pictures could also not tell the whole story as one is more blown up. Really youd have to blow up the pictures to tell.

                  That being said i would love to see either monitor reviewed here if thats possible.

                  #69392
                  PCM2

                    Yeah, the camera settings are definitely odd there on the Reddit thread and you’re right about the other variables in play you can’t be sure about. I would also add unknown warmup period to that list, because VA models like this are particularly temperature sensitive and don’t perform their best straight away. There is no way the dark background should look that dark if the camera is set up ‘correctly’. I am extremely careful to set up the camera in a way that captures the motion in a very similar way to what the eye would see, with emphasis on any trailing looking as close ‘to the eye’ as possible. I do plan to review the Samsung Odyssey Neo G7 shortly, which is really the model I favour out of the two ~32″ Odyssey Neo models. So that should give some good insight into motion performance of that one.

                    #69394
                    Gigamike

                      Great! Its the one of the two that i would love to use due to the anti glare coating being so light. Motion clarity is also so important to me aswell and its so hard to tell what that looks like when all HWUnboxed does is favor the pixel timings over the actual comprehensive motion images that youd physically see. The G8 looks like it has “better motion” across the refresh range, but i cant be sure. OFten the 240hz looks great at the high and but suffers at the lower end. Maybe thats not the case here.

                      Looking forward to the review! Thanks

                      #69410
                      dynastes

                        Thank you for summarizing all of this. Looking forward to your review of the Neo G7 as well. Seems to be the better monitor for me personally, owing to the lighter coating, no scanlines, the lower price and still good performance characteristics (which you will hopefully confirm, too).

                        There is a third monitor in this 32″-VA-UHD segment, which I still think is being somewhat widely disregarded: Gigabyte’s M32UC. Sure, that one offers no Mini-LED FALD and therefore basically no HDR capabilities. Otherwise, though, it could be an interesting alternative (again lower price, less aggressive curve, KVM switch). Rtings has reviewed it somewhat positively, but tftcentral, HUB and pcmonitors, which I personally hold in higher esteem, have not posted anything on it. @PCM2, do you have anything to share about that model by any chance?

                        #69418
                        PCM2

                          You’re welcome, dynastes.

                          The main potential issue I’m aware of with the M32UC is the pixel responsiveness – it simply isn’t on par with the Samsung models, even though it is still quite decent for a VA model. This is reflected to an extent by the values (perhaps more tellingly the pursuit photos) provided by RTINGS, though they only look at a fairly limited range of transitions so don’t tell the full story. I’ve also received feedback from somebody who purchased both the M32UC and M32U to perform some side by side testing and aiming to keep the one they preferred. Although they’re generally a fan of VA and did enjoy that aspect of the ‘UC’, they simply didn’t didn’t find it responsive enough especially when gaming. They felt it was clearly inferior when comparing side by side with the ‘U’ model. It’s very subjective of course and it’s at a level where some people would be just fine with it – there isn’t the sort of extensive ‘smeary’ trailing VA models typically have. Although it does have a lot of what I’d describe as ‘powdery’ trailing and some of this could be described as ‘smeary’ for some transitions involving certain dark shades. He also commented on text clarity being superior on the ‘U’ compared to the ‘UC’, though he wasn’t quite sure why – he thought it probably had something to do with the screen surface differences.

                          #69598
                          PCM2

                            Our review of the Samsung Odyssey Neo G7 has been published.

                            #69817
                            PCM2

                              I’d like to share another review of the Odyssey Neo G8, this time from Chris of TotallydubbedHD. His review includes some nice pursuit photographs which are shown at 240Hz where the video below should start. For the transitions shown it shows quite a bit of overshoot (in my view) even in the ‘Standard’ setting, which incidentally performs similarly to VRR enabled for pixel responsiveness here. The dark background shows this most clearly. It isn’t extreme overshoot and you can see this ramp up significantly at higher response time settings. There isn’t any real conventional trailing to speak of for the transitions shown, although the measurements provided in this review and elsewhere do confirm some isolated weaknesses where darker shades are involved. Basically similar to the sort of weaknesses observed on the Neo G7 at its maximum refresh rate, though not as widespread and replaced by more overshoot. Even the weakest transitions come nowhere close to giving the conventional VA ‘smeary’ trailing. At 120Hz both models perform fairly similarly (the ‘Console Gaming’ section of the below review includes some 120Hz pursuit photos).

                              Chris found it a capable monitor for competitive gaming in terms of pixel responsiveness and it’s always nice to have his take on that side of things – he has a lot of experience gaming competitively with a broad range of monitors. He didn’t find the overshoot problematic, obviously that is subjective as I always say. He also found the input lag agreeable with VRR disabled, which as he points out is generally preferred for competitive play anyway. He didn’t get on with the curve, particularly for competitive play where your eyes will scan between the centre and edges of the screen a lot. This is something I can completely relate to and also mentioned in my review of the Neo G7.

                            Viewing 13 posts - 1 through 13 (of 13 total)
                            • You must be logged in to reply to this topic.