It is currently 18 Jul 2024, 11:01

All times are UTC [ DST ]

Search found 8 matches

Author Message

 Jump to forum   Jump to topic

 Post subject: Sony HIL-* MUSE players
Posted: 19 Feb 2022, 13:16 

I was playing around with Sony HIL series players recently and wrote some notes on what i've found:

Maybe it could be interesting for other LD fans also.

 Jump to forum   Jump to topic

 Post subject: Re: Sony HIL-* MUSE players
Posted: 09 Mar 2022, 17:54 

Fantastic information! Thank you for creating this and sharing it here.

Have you thought about doing a deep dive into the KHS-160A? It seems like a lot was simplified when compared to KHS-140A.

Thank you for checking it out!
Regarding the KHS-160A, yes, thank you for your suggestion, it is on my ever-growing TODO list... :) Indeed, the 160A is much more compact!
Right now an HLD-1000 and a Toshiba XR-HD1 are on my desk.

 Jump to forum   Jump to topic

 Post subject: Re: Sony HIL-* MUSE players
Posted: 28 Aug 2022, 07:13 

Yes, very interesting and in-depth. Really looking forward to more of this :thumbup:

Did you decide to research this very specific topic out of curiosity or for a scientific project?

Thanks :-)
It was just curiosity now. I've did a small update on the docs though, but seems like i don't have the sophisticated equipment / parts / more time to spend on this project/topic. Anyway, it was interesting, i've learned what i wanted to learn and also had some fun while doing it.
Still, one future plan i have is to reverse engineer the code i've dumped from its MCUs to see what the system control software exactly does. And also the sw from a Pioneer HLD-1000 and a Toshiba clone. One of the MCUs in the Sony HIL-C1/1000 is very similar to the main MCU in the Pioneer HLD-1000, it seems both are MELPS740 arch (probably HLD-X9, HLD-X0 and many other Pioneer players too, but i don't have access to an X9/X0). Then it would be possible to create custom software (firmware, as we would call it today), or maybe replace the MCU itself with a modern one, which could be useful for testing/debugging the player, gathering "secret info" :lol: , creating custom OSD/front panel, use different remote control protocols, interface with a PC, etc . Well, not much practical consumer use, mostly only for fun.

 Jump to forum   Jump to topic

Posted: 21 Sep 2022, 16:30 

I was investigating that possibility too.

Very short answer: no. :(

Short answer: yes, but doesn't worth it. :eh:

Long answer: depends on the device what you can extract from it. :roll:
For example, the Sony SAU-500MN has 3 three main ICs. (And SAU-300MN too - but i don't have that one to confirm, but these two seem very very similar, if not the same).
Simplified: one for MUSE audio, one for MUSE processing, one for HD->NTSC format conversion (zoom/wide/full) and downscaling (1 NTSC line is produced using 5 neighboring lines from the MUSE signal).
From the MUSE signal processor IC (CXD2020) to the NTSC format conversion IC (CXD2021) the luminance and chrominance signals are transmitted via two 8-bit digital channels. This one chroma signal contains both B/R multiplexed. So you would need to build some custom signal processing to make it into a HD component signal. It is not impossible, but requires probably lots of effort and additional devices. Doesn't really worth it unless you enjoy hacking with it, it's much easier to get a MUSE decoder with HD component output. Also the "quality of decoding" is lower in M-N converters, because in order to save costs some image processing were omitted (e.g. anti-aliasing, and interpolation only in the same field to save VRAM, possibly losing some horizontal resolution too), which may be OK for the final NTSC output on an SD display, but maybe noticeable when using a HD signal with HD display. So even if you manage to convert the Y+C output of the CXD2020 into something useful, you'll possibly get a lower quality (due to the lack of some post processing/filtering) image than from a "real" MUSE decoder.
Possibly this is (partly) true for M-N converters made by other manufacturers too.

An easy thing to try is to get the SD component signal before the NTSC encoder IC (CXA1219). It's still SD but can be a bit better than NTSC. Maybe also doesn't worth hacking with it.

One thing that would be interesting to me: Is it possible to get the HD component signal out of a MUSE - NTSC Converter? I have one but of course I would like to get a HD signal (once I find a Hi-Vision LD for a decent price). If it is a two-step process - first decoding, then downsampling - it could be possible.

 Jump to forum   Jump to topic

Posted: 11 Feb 2023, 16:05 

I was experimenting with the RS422 system control port on the Panasonic TU-AHD100 , and made a simple util to control it from a PC.
Well, it works for me, and thought others may find it useful or at least fun. Of course you'll need some simple HW to interface your PC with RS422.

Something like this:

You can find it on github, here .

 Jump to forum   Jump to topic

 Post subject: Re: Sony MST-1000
Posted: 01 Apr 2023, 06:16 

Indeed, it would look great beside a HIL-C1!

Based on the Yauc photos and other info, this looks very similar to the 1st generation of the MUSE LSI chipset.
Sure it has many ASICs, but as the name states it's already Large Scale Integration :)
The "original" prototype MUSE decoder consisted of around 3000 discrete ICs with 43 PCBs in a rack, and consumed ~1500W (and probably required at least two persons to move it).

The MUSE signal processing is done in the digital domain, but still, not all decoders do everything the same way of course.
The main focus of the development was on reducing the cost/size/power consumption-heat dissipation of MUSE decoders, but this doesn't mean that (all) newer decoders sacrificed quality over cost savings.
Characteristics of the filters used for interpolation can vary, which can make a difference.
And possibly different methods for motion detection were implemented in different MUSE VP chip(set)s.
Also, some decoders use 8-bit, others use 10-bit ADC/DAC circuits. MST-1000 should have 10-bit, but for MUSE LD players, that doesn't make any difference.
Many newer decoders can also contain additional video processing circuits to improve the image in their own way (e.g. "MUSE AI" feature in the Panasonic TU-AHD100(N)).
Fujitsu also developed a one-chip MUSE processor in 1995 (calling it 3rd generation MUSE LSI), stating that by using 7x7 filter instead of 7x5 (as in the 2.5th and possibly previous generation LSIs, maybe the prototype MUSE decoder used 5x4 for the 2D filter) significantly reduced ringing and improved horizontal resolution (400->480) in moving areas.
This also employed a 3x3 median filter instead of the previously used 3 tap high pass filter for noise reduction.

So, my guess is that the picture of MST-1000 shouldn't be bad, but newer decoders probably possess more "magic" features and can get more out of the input signal (probably mostly visible with moving areas).
But of course image quality is mainly subjective.

Definitely (at least to my mind) a much nicer looking unit cosmetically than the MSC3000/4000, I wonder how its video performance compares to these later units though, and especially bearing in mind the high quality innards that you note above Julien?

 Jump to forum   Jump to topic

 Post subject: Re: [MDP-850D] Crosstalk
Posted: 04 Sep 2023, 07:30 

I have two of these back at home. As others already stated, it's hard to tell from these photos, but probably what you got is normal.
I don't see any crosstalk either.

I remember experimenting by turning some pots on the main PCB, didn't really help with anything.
If you want to see how crosstalk looks like, slightly adjust the TILT trims, but then don't forget to turn them back. (Not recommended if you're not sure what you're doing)

Actually, this is my first player, watched hundreds of discs, still perfectly working.
Well it's not a high-end player, certainly not the best picture, but i like it.

 Jump to forum   Jump to topic

 Post subject: Re: [HM-D101] Just received
Posted: 09 Dec 2023, 10:48 

So the HM-D101 does appear to be a step up on the 4000 during normal playback mode although not quite a night & day difference?

More of a step down! The HM-D101 when holding a paused frame for a few minutes started having weird pink or green dots added to the video. Stepping one frame clears it.

Will play with it some more over the weekend.


HM-D101 (and Hitachi HD-M20) has the Fujitsu MB861xx series chipset, so basically it's the same as (or very very similar to) a Sony MST-2000.
Or to Sony MSC-3000, which also shares the same chipset except the "time base" / "output processor" parts were replaced (upgraded? :-) ) by Sony developed ones, others stayed the same.
The DAC and ADC are also made by Sony in the HM-D101, video DAC at the end is the exact same as in MSC-3000/4000 (and maybe MST-2000 too).
MSC-4000 should have more advanced digital filters (nicer interpolation).
Page 1 of 1 [ Search found 8 matches ]

All times are UTC [ DST ]

Jump to: