[Buyer’s Guide] Kaleidescape Quality Compared to Blu-ray & other Digital Sources

[DieZwei]

Isn’t it a display/tv/monitor VS projector thing?

Why would that be case?

 

[Movielover76]

Hello all,

It's been about a week and a half since my last update, and there is much progress to share.

In my last update to this thread, I noted that we had identified two underlying causes of the "block artifacts" that have been reported, especially in bright areas of HDR content. We wanted to have a little bit more certainty around those before sharing details, but now is the time.

As I think everyone reading this thread knows, Strato V, M, and E are based around a new chipset compared to the older-generation Strato players like Strato C, S, and the original Strato. This new chip has lots of great features, like the ability to play Dolby Vision content, decode multichannel audio, and so on. It also includes a video scaling engine with lots of capabilities including image enhancement functions that are used when up-scaling content (e.g. HD played out as 4K).

Useful as those image processing features are when upscaling, you don't want them on when playing content at its native size, because they can introduce artifacts. So of course we turn them off whenever we're not scaling.

Well, it turns out that the chip had a sneaky trick up its sleeve. Readers of this thread will also know that as part of playback of letterboxed content, we blank out the part of the image that is the black letterbox bars. This is so that they appear truly black during playback, which looks better and also facilitates the function of things like auto-aspect detection in video processors. It turns out that for the chip's internal purposes, "blank out these portions of the image" is a feature of the scaler. Which means that although we have image processing turned off when the scaler is inactive, the scaler is in fact active much more often than we expected.

So, this is good news. We have made image comparisons now between the currently shipping software and an engineering build with all of the image-processing functionality disabled, and many of the "block artifacts" disappear. The image processing features also include some noise reduction that can cause a loss of grain detail, so turning those features off also restores that detail. That means that even without updating encoding, the playback of existing files will look better once we have deployed a fix to the player software to disable the unintended processing.

But, we're also making updates to our encoding. Without wading too far into the weeds on this, we have identified some improvements that can be made to our encoding "recipe" that make meaningful differences in preserving grain structure and minimizing blocky artifacts in the chroma channels of the encoded image. When combined with disablement of the inadvertent image processing, the resulting playback is significantly improved. Even on older Strato hardware that is not affected by the image processing bug, the updated encodes will deliver better quality.

Finally, an update on tools. I mentioned previously that we were working on improvements to our workflow and tools. In addition to improving our processes to better catch things like bogus HDR metadata supplied with the master files, we've also been working to improve our automated analysis of encoded movies. We already had automated tools that were run on every movie to ensure that the encoded files are preserving image detail. To that existing tool, we are now deploying a color analysis tool that will ensure that the encoded file's colors are faithful to the original. Color shifts like those that have been identified in this thread would immediately be flagged by this automation, so that corrective action can be taken. With our combined adjustments to both process and recipe, all known sources of color shift have been addressed, but this tool will provide a backstop to help ensure that we catch some future source of error. (Note: we are still looking at reported difference in red saturation on a few recent titles.)

Now, having said all of this, the obvious question is: when will we start seeing these improvements?

On the encoding front, we have already begun using an updated recipe on some recent releases with good results, so we will continue to use it on suitable content even as we continue to refine it. We will also continue making updates to titles that were affected by color shift or metadata issues, and those will naturally be processed with the latest recipes as well.

On the player side we are working on a software update that will disable the undesirable image processing. This turns out to be a bit of an invasive change, because we do need some of this functionality to be engaged when we're actually scaling content, which includes both upscaling (HD -> 4K) and certain CinemaScape modes. This makes a quick-fix change too risky to include in the next kOS software update, which is already nearing completion and getting ready for final testing. However, this work is being performed with a high priority, and we will release it in a subsequent update as soon as it has been adequately tested. We'll be able to share more about that release timeframe once the work is done and in testing.

I hope that this update will reassure you all that we are taking any and all quality issues very seriously, and the work to resolve them is being done at the highest priority.

Let me close by just making a note here that I will be on vacation for the next week, so please don't be alarmed if I'm not especially responsive here on the forums. I'll try to look in when I can, and I know John will also be watching (although please note that next Monday is also a holiday for him in Canada).

Thanks for reading this far. I'll share more updates as I have them.

Many thanks, Mike. Given that this issue, in particular, is mostly down to an unexpected side-effect in newer hardware, I would hope those who were calling for unnamed and "guilty without trial" individuals to be sacked could calm down a bit, for everyone's sanity.

 

[cinelife]

OK, so that means the issue is introduced by the Strato V @Rmarci is using for his checks. (it´s not that the question has been raised already if it might be the player...)
That explains why the majority (incl. me and most of our clients) is not seeing blocking issues and the like - they are using Strato C/S players.

Agree. I stopped posting in this thread because my "pushing back" was seen as a biased dealer response, but from the beginning I was not seeing blocking in any of the movies I viewed, including several mentioned in this thread, and frankly the quality of the video presented was excellent to my eyes. I do however recognize that while I watch movies for the story being presented, there are some people the watch with the intent of looking for imperfections, and if that's more important to them, then I respect that right and accept that we watch movies for different reasons.

As to the player contributing to some of the issues discussed, I personally have no idea, but I do know I haven't seen any blocking on the Strato-C's I use in 3 systems, nor in the one V I still use on occasion in standalone mode.

Jim

 

[Rmarci]

Thanks for the update, but at this point, I’ll believe it when I see it. I’m having a hard time trusting any update from them.

many of the "block artifacts" disappear.

Not all of them???

Isn’t it a display/tv/monitor VS projector thing?

???
The capture card doesn't care about the display...

I do know I haven't seen any blocking on the Strato-C's I use in 3 systems, nor in the one V I still use on occasion in standalone mode.

ok

I would hope those who were calling for unnamed and "guilty without trial" individuals

Huh, still half of their content has a green or red tint and is poorly encoded with obvious DNR and artificially boosted brightness... not to mention the red push again on the new release Caught Stealing, even after they supposedly implemented color checks.

 

[mariyoshikong]

If I read this correctly, he is attributing the “DNR” and macroblocking (which I have never seen on my Strato C) to the unintended scaling.

This also confirms something I checked last week with Superman where I compared @Rmarci screenshots with both K and Disc. The screenshots didn’t match what I was seeing on either K or Disc.

 

[Rmarci]

to the unintended scaling.

Some of it, maybe, but I can assure you that they were doing DNR and adding a fake grain overlay on many of their encodes. Just look at the KingsMan, iT-2017 and Pixels caps.

 

[mariyoshikong]

Some of it, maybe, but I can assure you that they were doing DNR and adding a fake grain overlay on many of their encodes. Just look at the KingsMan, iT-2017 and Pixels caps.

I will compare It in a little bit to see how that looks as far as DNR goes.

 

[Furnace Inferno]

I did wonder if it was the new chips in these players when thinking about the A.I scaling options it has in the setup menus and whether they were always active at the base hardware level because other media players and displays run into the same issue. They need to realise they aren’t doing anyone a favour by messing with the image!

Well thanks for the update @MikeKobb. I will be removing myself from the thread now we know what the issue is since I have a very different perspective on what constitutes as a major issue that would prevent me buying movies, I think I’ve spend $700 in the last month alone, yes I have a problem .

 

[Rmarci]

I love how now everybody is 100% blaming the player when Mike clearly stated that the block artifacts were caused by a combination of two things, and that the player was only making the issue more obvious. In other words, their encoding settings or workflow are introducing the artifacts, and the player is just making them more visible but this wouldn’t even be an issue if they hadn’t introduced the artifacts in their poor encoding in the first place.


quote from mike:
''many of the "block artifacts" disappear.''
'' These are not technically “macroblock” artifacts, but they do look similar. The cause appears to be a combination of two factors, and the good news is that we are testing a fix for one of them already. We’re still working on the other one.''

 

[mariyoshikong]

I downloaded It 2017 on my Terra and compared it with the UHD disc. I can see the red push on the Kaleidescape version but I am not really seeing anything that looks like DNR with my Strato C. I compared freeze frame and in motion and didn’t notice any real difference in fine details. I might hook the Strato up to my monitor later to do a better comparison.

 

[Rmarci]

That's because you didn't read my post about it.

Post in thread '[Buyer’s Guide] Kaleidescape Quality Compared to Blu-ray & other Digital Sources'

Sep 19, 2025

It’s just the opening shot that has this strong DNR, so I assume there was something in the master they wanted to remove.
I’ll have to check my UHD Blu-ray to see if I can spot anything.

For example, in IT (2017), the UHD Blu-ray has with some weird noise (black dots) that last for a few frames at 1:10:00. KS used strong DNR to remove the noise but also stripped away all the detail from the shots, then added a heavy fake grain overlay. Personally, I much prefer the UHD Blu-ray without any DNR at all, even with its defects.

It 2017 KSCAPE vs UHDBD | Slowpoke Pics

Slowpoke Pics image comparison service

slow.pics

They did something similar...

• Rmarci

• 

 

[mariyoshikong]

That's because you didn't read my post about it.

Post in thread '[Buyer’s Guide] Kaleidescape Quality Compared to Blu-ray & other Digital Sources'

Sep 19, 2025

It’s just the opening shot that has this strong DNR, so I assume there was something in the master they wanted to remove.
I’ll have to check my UHD Blu-ray to see if I can spot anything.

For example, in IT (2017), the UHD Blu-ray has with some weird noise (black dots) that last for a few frames at 1:10:00. KS used strong DNR to remove the noise but also stripped away all the detail from the shots, then added a heavy fake grain overlay. Personally, I much prefer the UHD Blu-ray without any DNR at all, even with its defects.

It 2017 KSCAPE vs UHDBD | Slowpoke Pics

Slowpoke Pics image comparison service

slow.pics

They did something similar...

• Rmarci

• 

I did. Everything looks pretty much identical to me except for the red push. Again, I will check on my monitor later to confirm.

 

[MikeKobb]

I love how now everybody is 100% blaming the player when Mike clearly stated that the block artifacts were caused by a combination of two things, and that the player was only making the issue more obvious. In other words, their encoding settings or workflow are introducing the artifacts, and the player is just making them more visible but this wouldn’t even be an issue if they hadn’t introduced the artifacts in their poor encoding in the first place.


quote from mike:
''many of the "block artifacts" disappear.''
'' These are not technically “macroblock” artifacts, but they do look similar. The cause appears to be a combination of two factors, and the good news is that we are testing a fix for one of them already. We’re still working on the other one.''

Let me clarify a bit.

Before I do, I would ask everyone to keep in mind that the content reviewed in this thread was processed over a span of nearly a decade. As I’ve already acknowledged, some of the older encodes (Pixels being an excellent example) were not up to our current standard of quality, and were already scheduled to be revisited and updated.

Now, moving on to the clarifications. I wish that I could illustrate the following with images, but as I think I explained previously, we can’t extract frames from copyrighted content and post them, because those frames are not ours to show. I’ll do my best to describe, though. Maybe at some point we can get some frame grabs from open source test content to share.

Most of the block artifacts are in fact introduced directly by the unintended image processing in the player. With that processing disabled, most of them disappear entirely. What we still see with the processing off is that on some content there can be some non-uniformity in the grain structure, which while not strictly a “block artifact” does produce a similar visual effect. And, in some content there is some blockiness or banding in the chroma channels. Addressing these remaining aspects is where the recipe updates come in. The updated recipe that we’ve begun to introduce does a good job, especially in improving that non-uniform grain structure. We’re continuing to refine it to steer a bit more of the available bandwidth to chroma to fully address what we see there.

 

[Rmarci]

Most of the block artifacts are in fact introduced directly by the unintended image processing in the player.

When can we expect a KS OS update? weeks or months? Considering this is a major update, and the fact that we still haven’t gotten the fix for the over-crop bug, I really hope it won’t take months to get my $5K CAD device to work properly, something my $50 Firestick does correctly.

I did. Everything looks pretty much identical to me except for the red push. Again, I will check on my monitor later to confirm.

Well, look at the screencaps . If the player was introducing that much DNR, all my other caps would look just as bad.
Or maybe that title got updated?

 

[MikeKobb]

When can we expect a KS OS update? weeks or months?

As I said above, it won’t be in the next kOS update because it’s too late in the release cycle to implement and adequately test the necessary changes. While it is being actively worked on, we don’t forecast release dates for unfinished work because it’s always possible that some unanticipated problem crops up that causes a delay. The best I can do is to say, “as soon as it’s implemented and tested.”

 

[Rmarci]

Well, I didn’t ask for a specific date, just whether it’s a matter of weeks or months?

 

[MikeKobb]

I can say confidently that it will be more than a month, since it has to follow the upcoming kOS release that isn’t ready to start deployment yet. Beyond that, I really can’t be more specific for the reasons described above. The aim is definitely to get it out to people as soon as we can.

 

[Trevor]

I know you can't answer Mike, but what's in this next kOS update that's so much more important than fixing the identified image quality issues?

Halt the rollout of this (likely) minor update and roll those changes in with the changes we actually need. If that gets the image quality update out even a week earlier than that's what you guys should be doing.

Otherwise, digging the detective work you guys have put in. Sounds like the team is on the case and good things will come.

 

[mariyoshikong]

Well, look at the screencaps . If the player was introducing that much DNR, all my other caps would look just as bad.
Or maybe that title got updated?

I just compared these on my QD-OLED monitor. The Kaleidescape version on my Strato C looks pretty close to the UHD disc version in your comparison screenshot (and also an MKV of the disc version playing back in JRiver) for the scene where they are walking down the hill with the train in the background. I don't see the grain in the clouds like your screenshot shows and I don't even see a red push on my monitor which makes me wonder if I have something messed up in my theater or my brain was making stuff up with regards to that.

It is possible that the Kaleidescape version is a little softer which might indicate DNR but unfortunately the monitor has to scale this down so it is hard to say definitively.

I wonder if this title got an update at some point.

 

[Rmarci]

for the scene where they are walking down the hill with the train in the background

Wait, what? That’s not the DNRed shot I was talking about in the post you said you read (obviously you didn't read it)...

This is 100% DNR + fake grain overlay