Epson LS12000 Dynamic SDR + HDR – Take 2

I spent more time with the LS12000 today and there are a few more findings, settings and a present for you in this article.

CCSS

30 Jun 2022: This section has been edited for accuracy.

Colorimeter Calibration Spectral Set (CCSS) is the file ArgyllCMS and HCFR (current release as of 30 Jun ’22 here) uses to tailor colorimeters to displays that support CCSS files. For example, the i1 Display Pro is calibrated in the factory and this calibration is stored in firmware. The i1 Display Pro also comes with EDR files (same as a CCSS but in a different file format) that allows it to be tailored to the display technology you want to calibrate. The issue with the EDR set is that they haven’t been updated for ages and they are not really suitable for laser projectors.

While it would be more accurate to create a CCMX for a particular display – colorimeter – spectrometer combination, a benefit of CCSS is that it is transferrable between display units and between colorimeters. The CCMX is not really transferrable so it is best not to download one and use matrix offsets from someone else’s CCMX / correction matrix

To use the CCSS calculated for the LS12000, you need to install HCFR / ArgyllCMS and then pop it into a Color folder under C:\Users\YOUR_NAME\AppData\Roaming\Color. You will need to show hidden files and folders in Windows to be able to access this folder. Then you can use HCFR – or any other application that uses Argyll – to calibrate the LS12000 with your i1 Display Pro / Colorchecker Display / Colormuki Display / Spyder 4 / Spyder 5 devices. (I don’t actually think the Spyder X can use CCSS files, unfortunately.)

You can download CCSS file below for the Epson LS12000. The CCSS was taken with the laser at 466hours. It can be used between about 0 and 5000 hours on the laser after which it might lose it accuracy somewhat. The below CCSS was taken with an i1 Pro 3 in High Resolution (3nm) mode.

If you’d like to calibrate your Epson, you can find some resources here.

Please note that due to unit and screen variances, the calibration may not be transferrable. However, the calibration settings might help tame your LS12000 in dynamic mode until you get a professional calibrator, or you do a calibration yourself.

I also confess that this is still not complete victory in taming dynamic mode on the LS12000. It usually takes a few tries before I get the hang of such a non-linear mode. These modes are challenging to calibrate regardless of your experience level.

Dynamic Calibration – SDR Torch Mode

See the adjustments below. I used two-point greyscale with this calibration to demonstrate how such a thing could be achieved and it lined up relatively well. I started doing an 11-point greyscale but ran out of time – I wanted to focus on the CMS this time. You can use the previously published white balance and 11-point greyscale with the CMS adjustments if you want! That was more detailed but didn’t have the CMS adjustments. Here they are below.

  • Dynamic Picture Mode
  • Laser power: 100% (torch mode)
  • Dynamic Contrast: High Speed
  • Gamma: -1
  • Dynamic Gamma: 8
  • Image Enhancement –> dynamic contrast: 8

See White Balance and CMS adjustments below:

Notes

Calibrating dynamic mode on the Epson is a bit of a challenge to say the least. This is why most professional calibrators won’t even touch such a picture mode. Natural would have taken less time and been a lot more accurate. But what’s the fun in that – especially as dynamic mode has aggressive laser dimming.

The greyscale in SDR lined up nicely. See below. I could have done better by adjusting it using a 11-point greyscale but wanted to focus on the CMS.

The Colour Management System (CMS) does not switch into REC709 mode in Dynamic so the gamut is oversaturated, especially in red and green. However, it can be tamed to a point where it looks accurate under 75% saturations. See the before and after below.

Dynamic Calibration – HDR Torch Mode

  • Dynamic Picture Mode
  • Laser Power: 100%
  • Brightness: 2 (due to having touched offsets) – you can use an HDR Black Level adjustment pattern to set black to code level 64 (to standard ) or to 77 (to get around lifted black level in mastering) if this setting doesn’t quite work. You can also lower this to 1 if the blue push near black bothers you.
  • Dynamic Contrast: High Speed
  • Gamma: 0 ( -1 for dark cave might work better)
  • Dynamic Gamma: 8
  • Image Enhancement –> dynamic contrast: 8
  • Adjusting blue push near black: I didn’t have time for this today, but you could use the bottom stimulus level on the 11-point greyscale controls (stimulus 1) to increase green and red, then re-adjust the brightness control using an HDR Black Level pattern.

Notes

There are two methods to calibrating the Epson for HDR:

  1. Calibrate to 2.2 gamma in SDR – then check that EOTF lines up once switching the unit into HDR mode. It normally does!
  2. Use an EOTF calibration on top of the SDR 2.2 gamma cal and change the EOTF multiplier / brightness after switching Epson into HDR mode. I will publish such a calibration for the TW9400 / UB6050 / UB5050 in the next week or so to demonstrate this. It allows users with bigger screens to have more HDR brightness. It can be used with an LS12000 as a method also.

Thankfully, the EOTF was correct when the Epson was switched into HDR after having calibrated the gamma to 2.2. However, the greyscale did need more work specifically in HDR. The issue in HDR mode is that I introduced a slight blue push in the low end that is difficult to calibrate out without using multi-point greyscale controls. That is certainly a better strategy.

As mentioned during my first calibration, the issue with the new 11-point controls is that they don’t line up 100% with the expected stimulus levels (at least not in dynamic mode) so there’s quite a bit of playing around. The onboard patterns didn’t seem to be correctly lined up either or didn’t respond to calibration controls properly. At least the UB6050’s controls didn’t have this issue. It could be a bug. In any case, use external patterns and going from the top: 11 is meant to be 100% stimulus, 10 is 90% stimulus, 9 is 80% stimulus and so on… at least in SDR!

I would need more time with the unit to figure this out. One way to check the stimulus levels is to do a 30-40 point greyscale and then jack up one colour at a time per stimulus then re-profile greyscale to see where the colour spike occurred. Then repeat with the next stimulus. Again, with more time I could have done this.

Greyscale: this is only serviceable. But the blue push in the low end was better than having a red push higher up affecting skin tones. So this can be tightened up massively using the 11-point greyscale if you don’t mind fiddling around with it. In which case, you should NOT use any offsets and use only gains then work down the 11-point greyscale.

EOTF lined up well (tone mapped from 60% stimulus upwards – this changes per HDR slider setting)

CMS was quite challenging in HDR. I only remembered to take a picture of the gamut after CMS adjustment. I could have made it better with more time, but this is a fine first attempt especially because the gamut is not linear in dynamic at all. But it did improve the picture quite a bit and got rid of the ridiculously oversaturated colours. It is interesting that colour luminance was kind of alright in Dynamic this time, but the saturations were crazy. This is the opposite on the UB6050 / UB5050 / TW9400.

Conclusion

It normally takes me a couple of full days playing with a new display in such a non-linear picture mode to figure out the behaviour and then get the hang of it 100%. It doesn’t help that the controls aren’t yet bug-free in dynamic from what I can tell. Epson did improve HDR gamut and greyscale quite a bit on the UB6050 / 5050 / TW9400 using a firmware update. That unit had a bad red push near black which was difficult to calibrate out with earlier firmware. I would expect Epson to improve on the Dynamic mode this time around as well.

In any case, we enjoyed playing some SDR and HDR content on the Epson after the changes. The laser dimming in dynamic mode is quite aggressive and was fun to see.

23 thoughts on “Epson LS12000 Dynamic SDR + HDR – Take 2

Add yours

  1. When calibrating greyscale in HDR should I have Dynamic Contrast (laser dimming), Dynamic Gamma and
    Image Enhancement off or at 0? When then proceeding Calibrating ETOF, I know from reading your guide you set Image Enhancement (dynamic contrast / HDR slider) to 8. Do I turn back on Dynamic Contrast (laser dimming) or Dynamic Gamma when Calibrating ETOF?

    Like

    1. Hi Michael,
      Dynamic contrast, dynamic gamma and laser dimming must be turned off for the duration of the calibration as they are dynamic systems. I think you recall that in The Display Calibration Guide, these are called dynamic light control and dynamic processing and the prerequisites of each procedure in the guide are “additive”. You don’t turn anything back on unless you’re told in the next procedure. Otherwise only at the very end of the guide.

      However, the HDR slider is not a dynamic system, it’s manually setting your HDR clipping point and multiplier (so your EOTF tracking). If you calibrate to gamma 2.2 in SDR, then the EOTF will line up correctly across all slider positions, only the multiplier and clipping point will keep changing.

      However, it is easiest to check this at slider position 8 when feeding the unit HDR patterns.

      I would recommend for everyone to start with SDR calibration though as you can improve the Epson massively by calibrating greyscale and gamma in SDR. It will also affect HDR correctly. Then if you wanted to save a separate memory to tighten up the gamut for HDR, you can and you have a fully calibrated HDR mode due to the geniuses at Epson.

      In the next round, you can then try greyscale – EOTF – Gamut fully end to end in HDR. It won’t be nearly as easy but now you have more experience, a point of reference of what looks correct – and some sense of success – to help you through it.

      I hope that makes sense. 🙂

      Like

      1. Thank you for you help. That does help clear it up.
        I have calibrated SDR already was just in sure about whether have those extra settings if when calibrating HDR on a Projector. I am use to doing plasma, led and oled flat panel screens they don’t have these in HDR as they have dynamic tone mapping . So dealing with static controls and having to calibrate ETOF is new to me.

        Like

      2. Ok, that makes more sense now.

        Have you decided what brightness you’ll be using?

        It really isn’t recommended to use 100% laser for SDR as it’s way too bright. The reason it was used for this calibration was because the store it was done for had a massive screen and wanted all the brightness they could get for demos.

        If it was for home use, I’d just use 30-40% laser for SDR and maybe 70% for HDR in a dedicated theatre. In a living room with white walls, I might go higher, especially if there’s some ambient light in the room or I’d use it for gaming where I like a bigger impact.

        Hence I said it’s best to follow The Display Calibration Guide and figure it out for your own setup and preferences.

        Like

      3. You also asked me about ALR screens. The only thing you need to be aware of is with the incidence of reflection. For example, a retro-reflective ALR will behave differently to an angular-reflective screen.

        However, the setup is the same in all cases: You need to adjust the sensor in a way that gets you the highest brightness reads on a 100% white pattern.

        For ALR screens, look at the manual of the screen and make sure to point the sensor into the reflected light by working out the reflection angle based on projector location and the type of reflectivity the screen has (from the manual).

        I hope that helps.

        Like

  2. Thank you, that helps. I have my projector mounted so that angle is slightly less than 5% from the lens to the center of the screen. I will adjust the sensor so the is receiving light from its mirror angle from its distance from the screen. I could also get a few readings from a few heights close to where that should be to find the brightest point.

    I plan on getting a HD Fury Vertex 2 sometime in the winter to put between my receiver and project to output LLDV. Should I re-calibrate HDR with it connected? or should I be using the HDR calibration I had already will have done with out it connected as a baseline for output conversion to LLDV?

    Like

    1. Nope, just use the calibration you have done for HDR with the slider at position 8. However, for darker content you can go upto 4 (those bastardized Dolby Vision streams where they put SDR into a DV container).

      There’s a bit more nuanced information about DV and LLDV in Volume 2 which I’d recommend reading if you want to learn more.

      Like

  3. Quick question, I did not see any custom gamma controls on the Epson LS12000, so I would be left to profile gamma using the 11 step gray scale by increasing all colors in that step equally, correct? As messing with contrast adjust gamma is not optimal. I would be trying to correct gamma using grey scale without throwing the gray scale calibration too much out of balance?

    Like

    1. The LS12000 has both, Michael. But I prefer to adjust gamma using multi-point greyscale on the Epson to retain the gamma selection. Yes, you would need to change colours equally (which might translate to different amount of clicks per RGB channel) as you measure a particular patch and monitoring the greyscale as well as the Y (luminance) target.
      The Display Calibration Guide has an exact procedure you can follow. Thx!

      Like

  4. You mention in your guide that generating patterns from pc can have a problem with accuracy for HDR. Would this still be a issue when casting patterns from a windows 10 pc using HCFR to a Nvidia TV Shield with the pattern generator instead of outputting from the pc though HDMI. the windows 10 machine on has Intel UHD Graphics 620 graphics. In you guide for SDR you mentioned you could set black and white clipping (contrast and brightness) by profiling near black and near white patterns and adjusting the brightness and contrast until you got different (unique) reading across each 5 points. Is this method for adjusting brightness and contrast still applicable for HDR as well.

    Like

    1. Hi Michael, HDR pattern generation is a bitch but your intel graphics have a much better chance of being correct than the Nvidia Shield. As long as you can force your display into HDR, I would use HDMI out from your laptop. Remember than the patterns will be sent in an SDR container, don’t switch windows into HDR.

      Setting brightness in HDR is best done with the HDR black level pattern from a disc frankly. To ensure your laptop is outputting digital black correctly, set it up using SDR first. It will also be correct for HDR as digital black is the same across both.

      Like

      1. You need to read the LS12000 articles and copy the settings from the articles. Apologies.
        If you want to learn calibration with the Spyder X then there’s the Display Calibration Guide but that will take a lot of reading and time!

        Like

  5. Have you tried adjusting the Color Temp and/or GM correction when calibrating Dynamic mode? Others have measured the default color temp to be around 7500k in Dynamic mode. The high GM correction of 4, might account for the strong green push in the upper grey scale range that is difficult to calibrate out.

    Like

    1. Hey Michael,
      As far as I remember they lowered peak white hence I didn’t go that route.
      If you do calibrate it, change the temp and GM correction. If it lowers 100% white then you’re losing brightness. If it increases 100% white then you’re on the right track.
      Remember, this mode is done for maximum light output with decent colour. Of course if you can’t pull it in line enough then experiment. 🙂

      Like

      1. That makes sense. One other question that I was unsure if i was interpreting right from your guide was when calibrating CMS, when measuring 100% white first are you measuring 100% at 100% intensity (not 75% intensity) then switching over to 75% pattern intensity to profile color and adjust CMS?

        Like

      2. Not quite. If you look at the procedure and executing it in steps, you first set up HCFR for 75% intensity (or 50% for HDR, DLP projectors and OLEDs) and then start measuring the primary colours. In the primary colour chart there is a 100% white patch which is scaled based on your intensity setting. That’s the one you must measure because HCFR scales all other CMS measurements to that. Measuring 100% white in the greyscale set (which is also scaled btw) is not going to cut it as HCFR won’t take that into consideration. Hope that makes sense.

        Like

  6. I have had read the procedures list, You may want make an edit to the wording as I found it a little unclear. “You must measure reference white first to ensure luminescence measurements will be correctly calculated by the tool.” It if this line read ” You must measure reference white (100% white) first in the pattern intensity you have chosen to profile CMS, to ensure luminescence measurements will be correctly calculated by the tool. ” I think it would be best understood written this way. Remember that the person has just come profiling White at %100 pattern intensity in the previously when calibrating gray scale in the procedure before. It would be a good way to remind them that they need to re-profile white at what ever intensity patterns they intend to use to calibrate CMS.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create a website or blog at WordPress.com

Up ↑

%d bloggers like this: