TCL C935 Owners and Discussion Thread

Jurin

Established Member
Joined
Mar 5, 2019
Messages
146
Reaction score
156
Points
74
Age
40
Location
Tampere, Finland
1662656567699.png


Picked up a 75" C935 and thought I might create an owners and general discussion thread on it.

Links to useful posts in the thread:

Impression thus far (Firmware V469 an some comparisons to my old Sony XG9505):

Positives:
  • Local dimming control when it comes to blooming is verygood (On Local dimming: High). Best I've probably ever seen. No trailing or obvious blooming (on real content, test pattern do show some at off angles). Overall much better than the XG9505 especially on HDR which on the Sony is a hot mess.
  • Good native contrast at ~6700:1 D65 with peak at 120cd/m2. Black floor with LD Off 0.018cd/m2
  • Very good viewing angles for a VA. Also much better than the XG9505.
  • Snappy ui (after proper bootup, can be sluggish at first)
  • Very impactful brightness
  • Quite good reflection handling. Some rainbow reflections though so avoid light sources directly behind the screen. Better than the XG9505.
  • There is PWM but the frequency seems to be pretty high and the amplitude pretty low so it's fairly easy on the eyes. TCL usually has been using a static 960hz so this might be as well. Sadly have no tools to measure it other than my eyes (I'm very sensitive to PWM and motion stutter [read: OLED]) but haven't had any headaches, sore eyes, migraines or other symptoms. This is a huge plus for TCL.
  • Motion is very good. Good amount of granularity in the settings. Easy to hit a sweet spot between a proper film look yet no SOE. Better on low framerate content than any OLED imho.
  • Color accuracy OTB is mediocre (too red & blue) in Movie mode but minor tweaks to 2pt brings it to acceptable levels of dE (max ~1.3 average .9 in WB). After proper 20pt adjustment it's down to roughly 0.5dE - 0.7dE on average.
  • Screen uniformity on my unit is good. No obvious issues in content. Some slight tone difference across the screen and very mild DSE. Left side is cooler than the right. Noticeable in uniform fields but very very rarely in content (mostly just static news tickers etc.)
  • Price is fair. Paid 1999e on discount in Finland.

Negatives / Bug list
  • Banding. It's obvious the processing doesn't have enough precision to produce clean gradients when compared to the Sony. Especially on darker tones.This is content specific though as most 10 bit content is pretty clean vs. 8 bit (thus also HDR vs. SDR). The decontouring filter doesn't do much to help and causes artefacts noticeable especially on moving faces.
    • Update: Might have been a bit quick to judge this based on playing youtube demo material as it uses an incorrect colour space setting. Most likely this is the reason for banding. Banding is much reduced on sources that use the correct color space. More on that later down the list.
  • Calibration is not as smooth as it is with the Sony. Partly because there's no Autocal (it's coming according to interview with Marek Maciejewski) and because the WB-adjustment doesn't really seems to have enough precision. +/- 1 can cause significant dE-changes especially on the darker end.
  • Overall processing not as good as it is on the Sony. Sharpening has ringing and artefacting, same with the denoise processing. Much harder to get a good quality outcome from poor quality sources vs. the XG9505. It's essential to feed it good quality material.
  • Local Dimming: 'High' crushes some shadow detail (in attempt to minimize blooming probably). Local Dimming: 'Low' crushes less but causes the blacks to float.
  • Local Dimming: 'High' can cause visible brightness fluctuations between scene cuts with different APL. Mostly affects SDR.
  • BUGS. Mostly this seems to be a problem of settings not saving properly. "picture enhancement" features seem to re-enable randomly and the ones you do need to be enabled are suddenly disabled (calibrated WB-settings and local dimming setting)
    • Update: Doing a 'Picture reset' and only using the "Per input" mode seems to help a lot here.
  • Trying to make sense what the "Apply to all inputs" in PQ settings actually does. What does the TV consider to be an "input" I have no idea as it seems to apply different settings based on the incoming signal vs. not an app or physical HDMI-input. Would be very helpful if the UI clearly stated what the currently used input & signal is in the picture settings menu.
  • The internal Youtube app apparently uses an incorrect color space (Defaults to Native) and greatly oversaturates SDR content. HDR looks correct (or the Native color space is close enough to Rec. 2020 for the errors to not be visible).
    • There is no way to force a particular colour space so you could manually fix the issue.
    • Kodi sometimes initially starts playing in the wrong color space but if you go in to the Color space setting and switch it between Auto -> Native and back to Auto it will snap in to the correct Color space. This is SDR, HDR might perform differently.
    • You can use an alternative Youtube app called SmartTubeNext that correctly autodetects the color space
  • Dynamic tone mapping over tracks the EOTF = is too bright to be of any use if you care about image accuracy. Although it does mostly give better shadow details.
  • Apple Homekit keeps giving notification to setup it even though I've skipped the notification several times. Had to start the setup and disable it in the Homekit app to get rid off it.
  • If you switch off the TV Mic (physical switch under the TV) you get bright orange leds blasting at your eyes even though you have the power led setting set to "OFF" and Google Assistant Voice Control is "OFF".
  • "Quick start" setting seems to do nothing (unless you shutdown and restart in about a minute). Normal startup takes 30s - 60s.
  • After watching HDR content in Movie mode it sometimes doesn't default back to SDR brightness setting when the HDR content stops playing. If you restore it back to where it was it might not save and the next time you start the TV it will be 100% again.
  • 2pt, 20pt and color space calibration settings fail to save seemingly randomly for some inputs while it works for others
    • Update: Seemed to be related to the overall settings not saving issue. After reset has been mostly fine.
  • Happened once: Local dimming seems to enable / disable randomly independent of what setting is actually set. Was wondering why the picture looked so flat.. checked the LD settings. It was high. Switched it to Off and then back to High and LD was working again.
  • If you want to occasionally use the internal speakers but mostly use an external sound system via optical you can't easily mute the speakers without having the mute icon on screen all the time. You have to either set it to 1 (where it's mostly inaudible) or menu dive in to the sound settings and disable the internal speakers by selecting optical out only. Please make the mute icon disappear after a few seconds.
  • Started shutting off and restarting randomly and did that three times in a row. On the fourth cycle it freezed on the Google TV startup animation and refused to start past that point (waited 30 minutes but nothing). Repeatedly plugged it in and out of the outlet and for some reason 1 in maybe 10 attempts gave a successful startup. Immediately did a factory reset and seems to be fine for now... continued below
  • The boot looping started after having entered the 20pt WB settings (applied to all sources on Movie mode) and switched from HDMI to Cable. I repeated the behaviour just now. If you enter 20pt adjustments and try to switch inputs it will crash and reboot. If you happen to have the startup mode set to "previous input" this causes it to crash again and then freeze on bootup. No idea if it's certain values in the 20pt WB that causes this.
    • (Update: Did an NVM-reset and factory reset once again. Has not crashed since. Hoping this fixed it.)

Overall this has potential to be a real winner in regards to PQ but for me it all comes down to if I can live with the constant menu hunting, bugs, random restarts and disabling of "image enhancing features" that seems to turn on by themselves. Really do hope TCL can sort out the inconsistencies and deal with the bugs. If you're only going to use an external streamer or other source via HDMI it might save you a lot of trouble. At least you only have to hunt settings for one input :D

Ask me anything..

Edit 9.9.2022: Updated bug list
Edit 10.9.2022: Updated info on reboot & freeze issue
Edit: 12.9.2022 Updated info on reboot & freeze issues after NVM-reset. Appers to be fixed.
Edit: 13.9.2022 Updated info on settings not saving after NVM-reset. Appers to be fixed.
Edit: 13.9.2022 Updated contrast ratio test results
Edit: 14.9.2022 Updated info on the color space bug (appears to only be YouTube SDR content affected)
Edit: 25.9.2022 Updated info on local dimming etc.
 
Last edited:
Such a basic question but what’s it like watching sports? Football in particular? Does the banding or motion processing stand up to it or not?
 
Such a basic question but what’s it like watching sports? Football in particular? Does the banding or motion processing stand up to it or not?
No footie to test at the moment I'm afraid but did watch some WRC-rally and seemed just fine.

There's also a dedicated "Sports mode" but it's a horrid vivid-mode equivalent on steroids. Just use Movie mode and crank up the brightness & motion settings to your liking. Nothing to fault on sports but do note that apart from snooker I don't watch sports really.
 
View attachment 1749508

Just picked up a 75" C935 and thought I might create an owners and general discussion thread on it.

I've had it for two days now and still undecided if I want to keep it. Currently dialing it in and calibrating (Calman Home / HCFR + i1Display Pro). Coming from a 65" Sony XG9505 which I still have and can compare to.

Initial impression thus far (Firmware V463):

Positives:
  • Local dimming control is superb. Best I've probably ever seen. No trailing or obvious blooming (on real content, test pattern do show some at off angles). Overall much better than the XG9505 especially on HDR which on the Sony is a hot mess.
  • Good native contrast (initially measured around 5000:1 with peak at 120cd/m2, will have to remeasure to be sure)
  • Very good viewing angles for a VA. Also much better than the XG9505.
  • Snappy ui (after proper bootup, can be sluggish at first)
  • Very impactful brightness
  • Quite good reflection handling. Some rainbow reflections though so avoid light sources directly behind the screen. Better than the XG9505.
  • There is PWM but the frequency seems to be pretty high and the amplitude pretty low so it's fairly easy on the eyes. TCL usually has been using a static 960hz so this might be as well. Sadly have no tools to measure it other than my eyes (I'm very sensitive to PWM and motion stutter [read: OLED]).
  • Motion in general is very good. Good amount of granularity in the settings.
  • Color accuracy OTB is decent (barely; too red & blue) in Movie mode and minor tweaks to 2pt brings it to acceptable levels of dE (max ~2.2ish). Still working to get it calibrated properly with the 20pt.
  • Screen uniformity on my unit is good. No obvious issues in content.
  • Price is fair. Paid 1999e on discount in Finland.

Negatives
  • Banding. It's obvious the processing doesn't have enough precision to produce clean gradients when compared to the Sony. This is content specific though as most 10 bit content is pretty clean vs. 8 bit. The decontouring filter doesn't do much to help and causes artefacts noticeable especially on moving faces.
  • Calibration is not as smooth as it is with the Sony. Partly because there's no Autocal (even if some sources claim it's Calman Ready or it's coming.. I'll believe it when I see it) and secondly because the WB-adjustment doesn't really seems to have enough precision. +/- 1 can cause significant dE-changes.
  • Overall processing is no where near as good as it is on the Sony. Sharpening has ringing and artefacting, same with the denoise processing. Much harder to get a good quality outcome from poor quality sources vs. the XG9505.
  • Biggest problem: BUGS. The software is riddled with bugs to the extent that trying to make sense what effects what and how is borderline impossible. "picture enhancement" features seem to re-enable randomly on startup and the ones you do need to be enabled are suddenly disabled (calibrated WB-settings and local dimming setting!)
  • Trying to make sense what the "Apply to all inputs" in PQ settings actually does. What does the TV consider to be an "input" I have no idea as it seems to apply different settings based on the incoming signal vs. not an app or physical HDMI-input. Different random settings applied on every youtube video?! Have fun going thru your image settings and calibration before every piece of content you watch. How fun.
  • Youtube app apparently uses an incorrect color space and greatly oversaturates some content
  • Dynamic tone mapping over tracks the EOTF = is too bright to be of any use if you care about image accuracy

Overall this has potential to be a real winner but for me it all comes down to if I can live with the constant menu hunting and disabling of "image enhancing features" that seems to turn on by themselves. Really do hope TCL can sort out the inconsistencies and deal with the bugs. If you're only going to use an external streamer or other source via HDMI it might save you a lot of trouble. At least you only have to hunt one input :D

Ask me anything..
To to sum up, they need to sort the firmware out which will vastly improve the settings and overall user experience. If they can do that , this tv is a great/possibly the best all rounder ?
 
To to sum up, they need to sort the firmware out which will vastly improve the settings and overall user experience. If they can do that , this tv is a great/possibly the best all rounder ?
Yes that would greatly improve the overall experience. As it is currently it's a chore to get it set up right. But that being said it's doable with a bit of patience and a good source. The end result is some truly stunning images.

I'm considering buying an Apple TV 4K or Shield to use as a source. That would mitigate some of the issues. But perhaps would introduce some new ones.

I know posting pictures of HDR content is kind of pointless as the camera can't fully handle the dynamic range. The range the C935 has is just mesmerising. Avatar 2 trailer from HDR-X.

1662664801263.png
1662664839880.png

1662664890272.png
1662664907529.png

1662664933064.png
1662664971153.png
 
Any uniformity pictures to share? (low, mid brightness levels)
And what was HDR brightness levels (nits) in different box sizes (..4%, 9-10%, 25%, 50%, 100%)
And one more, that moving ball/square local dimming test, how much ld edge crawling was there in game mode..
I had 65" version few weeks and was thinking if 75" performs better?
 
Any uniformity pictures to share? (low, mid brightness levels)
And what was HDR brightness levels (nits) in different box sizes (..4%, 9-10%, 25%, 50%, 100%)
And one more, that moving ball/square local dimming test, how much ld edge crawling was there in game mode..
I had 65" version few weeks and was thinking if 75" performs better?
Hoping to have time this weekend for proper measurements and ld testing. But what I can say about looking at calibration test windows on LD set to 'High' there is some crawling and initial dimming around the edges (Movie mode, expecting Game mode to behave similarly).

Did a 20pt WB last night on top of the 2pt and got dE well below 1. Most around .5. Was looking real good for a moment there. That's when it went in to a reboot loop and finally froze on the Google TV startup animation and had to do a factory reset this morning. Good times!
 
Would you be able to test for 4:4:4 4k 120hz? as I couldn't find if could, and based on vincent (HDTV Test) review of the C835, the C835 can not display 4:4:4 at 4k 120hz
 
Would you be able to test for 4:4:4 4k 120hz? as I couldn't find if could, and based on vincent (HDTV Test) review of the C835, the C835 can not display 4:4:4 at 4k 120hz
I'll have a look if I have the energy. But my educated guess is that it will behave exactly the same as the C835 giving the same SOC.

Honestly I'm at my ropes end with the C935 and will be returning it. Still unsure if I'll replace it with another C935 just to verify that my unit was somehow defected or get something else (X95K or Z9K, or nothing at all).

Gutted because really wanted to like the C935. And the PQ can be outstanding but it's just impossible to live with.
 
Last edited:
I'll have a look if I have the energy. But my educated guess is that it will behave exactly the same as the C835 giving the same SOC.

Honestly I'm at my ropes end with the C935 and will be returning it. Still unsure if I'll replace it with another C935 just to verify that my unit was somehow defected or get something else (X95K or Z9K, or nothing at all).

Gutted because really wanted to like the C935. And the PQ can be outstanding but it's just impossible to live with.
It does share the same SoC as the C835, but it is said that the C935 does have a co-processor which made me believe that maybe it can.

It sucks that you will be returning it, but i cant blame you if it full of bugs. Hopefully it will get better with firmware updates (the same way as the X11 got better). As it is the only other option for miniLED that is done very good compared to the competition with the exception of the qn90b, sadly the qn90b doesn't track PQ EOTF accurately, even in filmmaker mode. (unlike the S95B)

This is the google tv version, I cant imagine what we get with roku tv in the US, the r655 (c835) roku TV experience is bad.
 
It does share the same SoC as the C835, but it is said that the C935 does have a co-processor which made me believe that maybe it can.

... with the exception of the qn90b, sadly the qn90b doesn't track PQ EOTF accurately, even in filmmaker mode. (unlike the S95B)
Can you confirm a source of information for the potential co-processor on C935? Have not heard that personally. My observations with poor gradients also would indicate that there is no co-processor. Or it's just buggy implementation (most likely incorrect color space detection on content).

Yeah, sadly any Samsung QLED is out of the question for me because of PWM & stutter. Also I'm really not a fan of Samsung as a company and their design decisions on their TVs (EOTF tracking, calibration cheating, PWM etc.)
 
Would you be able to test for 4:4:4 4k 120hz? as I couldn't find if could, and based on vincent (HDTV Test) review of the C835, the C835 can not display 4:4:4 at 4k 120hz
No 4:4:4 at 120hz. In fact no 4K 120hz at all available at least on my Win11 PC with RX580 4GB. 4K 60Hz is full 4:4:4.

Update: Of course RX580 doesn't support 4K 120Hz over HDMI 2.0 spec port so no wonder. Ignore this. 4K 120Hz might work just fine on a better & newer GPU with proper HDMI 2.1 spec support.
 
Last edited:
Would you be able to test for 4:4:4 4k 120hz? as I couldn't find if could, and based on vincent (HDTV Test) review of the C835, the C835 can not display 4:4:4 at 4k 120hz

Update to my previous test; Of course RX580 doesn't support 4K 120Hz over HDMI 2.0 spec port so no wonder it didn't work. Ignore this my bad. 4K 120Hz might work just fine on a better & newer GPU with proper HDMI 2.1 spec support. Sadly have no such GPU on hand to test.
 
Some measurement results on SDR with peak aligned to ~120cd/m2 and 2pt WB calibration to D65 in place.

Local dimming OFF (Window 10%, APL made no difference to results as expected):
Black floor 0.018cd/m2
Contrast ratio 6700:1 (Native)

Local dimming LOW (Window 10%, APL 18% in attempt to defeat LD fluctuation on measurement):
Black floor 0.008cd/m2
Contrast ratio 15000:1
LD on LOW was actually very good. Gamma tracking was damn near perfect. You could maybe even use this to calibrate with APL patterns. Good option for SDR If you can live with the somewhat floating blacks. For HDR probably not great.

Local dimming HIGH (Window 10%, APL 18% in attempt to defeat LD fluctuation on measurement):
Black floor 0.0cd/m2
Contrast ratio infinite:1
LD on HIGH was difficult to measure as there is some pretty dramatic brightness manipulation going on to mitigate blooming. It also drastically increases the peak light output so have to dial down the brightness setting quite a bit. Gamma tracking wasn't very linear as expected. For giggles I also measured the peak light output on a 2% window which was ~94cd/m2 so expecting small highlights to be dimmed down somewhat. Pretty typical on FALD.

Overall I'd say that the C935 would benefit from a "MEDIUM" LD-setting that would track more linearly than HIGH but wouldn't float blacks like "LOW". Of course easy for me to criticise :D LD algorithms are hideously difficult.
 
Last edited:
Update on the crashing and rebooting. Did an NVM-reset and a factory reset after that once again. Has not crashed since then so I'm hopeful that fixed it. Did the 20pt WB in Movie mode and after applying to all inputs it did not crash as it did before. 🤞 Also the initial setup felt much faster after this procedure.
 
Last edited:
If you don’t get any more crashes, is the TV good enough for you to keep?
 
If you don’t get any more crashes, is the TV good enough for you to keep?
That's definitely one deal breaker removed. Another one is the settings not saving properly but I haven't seen that as much after the NVM-reset either so still testing that one.

One big issue that still remains is the incorrect colour space in the internal apps. Might go and just buy a Shield TV 4K Pro or Apple TV 4K which hopefully resolves it. Realistically in this price point for a 75" there's really nothing that competes even factoring in the price of an external streamer.

LG GNED91 75" is currently 1500€ which might be an option; although I'm not a fan of IPS panels on TVs. Sony X95K, A80K, Z9K are all stupid money right now so not a real option until at least Black Friday. Overall still a bit on the fence but more positive atm.
 
We currently use a ROKU 4K stick for our apps. Am I right in thinking you’re looking at a similar way forward? At least until the bugs get fixed? I really want to like this TV as for size, features and price point it’s difficult to beat!
 
... Am I right in thinking you’re looking at a similar way forward? At least until the bugs get fixed? I really want to like this TV as for size, features and price point it’s difficult to beat!
Yes similar strategy atm. Treat it as a dumb panel; nothing more. Honestly this has been one of the most frustrating TVs to test (and I've tried MANY). Huge potential and so close to perfection but fails on many key things imho.

This is TCL asking big boy money for a TV but delivering a toddler experience at best. If anyone from TCL is reading this: Do better. Please. Thank you. I'm trying so hard to take you seriously so please take your customers seriously. I'm not even asking for world class picture processing, scaling or anything like that (I can work around that with external sources) but the most basic things like the settings I make to actually be saved. Thank you. Enough ranting :D
 
It is good to see that one of the problems has been fixed with a reboot and NVM reset; hopefully, the other problems will get resolved with firmware updates.

Thanks for the contrast testing; 5700:1 native contrast is pretty good compared to the competition (~3600:1 for QN90B and ~1500:1 for X95K); I wished it matched the C835, which got 7500:1 tested by vincent.

How were the post-calibration result after the 20wp?

Did you measure the brightness windows in HDR and SDR?
 
It is good to see that one of the problems has been fixed with a reboot and NVM reset; hopefully, the other problems will get resolved with firmware updates.

Thanks for the contrast testing; 5700:1 native contrast is pretty good compared to the competition (~3600:1 for QN90B and ~1500:1 for X95K); I wished it matched the C835, which got 7500:1 tested by vincent.

How were the post-calibration result after the 20wp?

Did you measure the brightness windows in HDR and SDR?
Hoping so as well!

Yes the native contrast is quite good. But it is interesting that it's lower than on the C835. I have theories why...

My guess is that it's due to a wider viewing angle compensation film on the C935 or that it's just a bit more aggressive on the 75" vs. 65" that Vincent tested. It does have surprisingly good viewing angles and this is exactly how I prefer it vs. having 0.005cd/m2 lower black level. The number of zones more than compensates for this. Well done TCL on this one.

Also the "OD5" panel might affect contrast ratio. The actual lcd-panel is a few mm behind the front panel. This might cause some light leakage as you're effectively measuring a larger area of the panel.

I'll post some 20pt graphs later today with brightness windows for HDR. SDR brightness windows are kind of pointless imho. SDR should be calibrated to 120cd/m2 and anything above that is just preference. I can say that it will get eye searingly bright. Brighter than anyone would ever need.
 
Last edited:
Good info here mate. Seems like a quality HDR TV but with a few niggles that need ironed out.

With regard the upscaling is there any way you can show that the Sony does a far better job upscaling - not sure if that is even possible to post a side-by-side or similar. If not is it really a big difference to the eye?
 
So after 20pt WB to D65 and 2.4 gamma. This was basically a numbers exercise on how low can you go. Will have to test thoroughly with real content for banding etc. Gamma slightly under tracks on the lower end probably because trying to hide compression artefacts and the overall banding. Didn't want to calibrate that out so left a tiny amount of undertrack in the low end as well. Will see how it looks on real content but I'm very happy with these numbers.

Keep in mind that I'm a DIY calibrator and my meter is not profiled so there might some errors baked in that I can't compensate for. Even so.. you could add 2.0dE and still be within perfectly reasonable results.

1663002876030.png
 

The latest video from AVForums

Is Home Theater DEAD in 2024?
Subscribe to our YouTube channel
Back
Top Bottom