S-Video to CVBS quality lossless cable/converter - exist?

[miro@hotmail]

Hi!

I have for capturing great device allowing me to capture in FullHD. To capture my TV I bought HDCP Decoder which has also CVBS (Composite) video input. There I connected my hobby computer output Commodore 128, but S-Video by reduction.

C128 output has resolution from 640×200 up to 800×600 - and problem is simple:
- text and pixels are correctly visible
- colors are mismatched and flickering (I mean same problem like connecting CGA to composite)

I'm searching for low-cost S-Video signal lossless quality calbe/converter to composite signal - exist any?

In case of capture device, it has three inputs: HDMI, VGA/DVI and Component
In case of HDCP Decoder has 2 inputs: HDMI and Composite

Capture device has connected Full-HD TV to HDMI, Commodore 128 output by VGA converter (there's problem with deinterlacing - it's flickering) and Component is free.

So I'm thinking about two solutions:

- use cable/converter written above (if exist) to HDCP decoder
or
- buy Composite/S-Video to RGB converter

For me is important price - if will be SVHS2CVBS cheaper, I'll buy it and still will have free RGB or if is cheaper CVBS/SVHS2RGB buy it but it gives me free only Composite input on HDCP Decoder.

BTW. HDCP Decoder scales up resolution of CVBS up to 1600×1200 pixels.

Thank you for all advices.

Miro

[Brad]

This sounds like it isn't a question about converting S-Video to CVBS as the title suggests, but rather how to take the non-standard output of an old computer and turn it into a signal that can be captured. You did the same thing with your previous topic.

To answer the title, no.

[miro@hotmail]

OK, doesn't exist.
This computer has two video subsystems - one is VIC in LCA (CVBS) and VDC (RGBI - same as CGA, connector too).
I have converter of RGBI to SVHS, also USB grabber, but this grabber works not good with interlaced modes, in case of VIC (CVBS) works with real interlace correctly.
So, I bought CGA2VGA adapter and tried it capture with PEXHDCAP VGA input - picture displays everything but it's not stable - joggle...
Then I connected not for capture S-Video converter to my HP LCD with S-Video input and there's no problem with displaying it perfectly.
So, when I bought for my HDTV HDCP decoder it has also CVBS input - I tried to connect SVHS output to CVBS output by reduction and picture is best as I got, but composite video is not enough if is compared to S-Video... for this was Q.
My PEXHDCAP has also RGB input (Component video) - so I was balancing about price.
I could exist and SVHS to CVBS converter with lossless quality it's cheaper than CVBS/SVHS to RGB converter.

Finally I go to buy CV/SV to RGB converter - problem is only one: price.
I'm from eastern Europe - Slovakia - converter's price is $80 and my month salary is 300 €...

This was real reason to start this discussion.

Also, thank you for explaining me that it's not possible and I have no choice.

Miro

[pandy]

Not sure what are you asking for - lossless S-Video to CVBS (composite) converter is capacitor approx 330 - 470pF - add C signal trough this capacitor to Y signal (it is lossless as it can be lossless).

Please describe in simple way what are you trying to do.
Side to this perhaps you should consider different capture solution - for example something based on ancient Bt878 however with correct drivers to deal with nonstandard video modes produced by older video sources. (Check DScaler software - it should be able to deal with lot of weird video modes).

Btw - video converters are in fact very cheap but if you are not electronic guy then perhaps you need to ask someone in SK about help. Check http://stores.ebay.com/njytouch/LCD-controller-/_i.html?_fsub=3585597018 - usually at output one of this board RGB, LVDS signal is available - issue is that acquisition board should be correctly programmed (and there is no good datasheet - but there is always REI possible).

[miro@hotmail]

I decided to buy CV/SV to RGB converter, thank you for link.

Miro

[pandy]

I decided to buy CV/SV to RGB converter, thank you for link.

OK, up to you but IMHO 80$ is to much for this kind of circuitry and you should search for some TV PAL/SECAM decoder solution or use for example DVD recorder as video converter.

[miro@hotmail]

All my devices to use CV/SV2RGB:

HDMI splitter - for displaying TV on LCD and plug into HDCP Decoder
HDCP Decoder - to plug HDMI into PEXHDCAP HDMI input to capture HDTV / it has also CVBS input (output by HDMI) - this I want to use to capture C128 VIC desktop
PEXHDCAP I/O:
HDMI input is connected to HDTV or CVBS of HDCP decoder
VGA/DVI input is connected to C128 VDC Desktop (not stable picture) or by VGA Selector to Videorecorder output or TV standard output (this CVBS is also free 'cause I'm capturing HDTV)
RGB input is free - here I want to use CV/SV2RGB - SV to capture C128 VDC desktop and free CVBS.

So, all these devices allows all my wishes and gives still free 2 CVBS inputs:
CVBS scaled up to 1600×1200 by CV/SV2VGA
CVBS scaled up to 1920×1080 by CV/SV2RGB

Not cheap but ideal - my task is only one - save $80 month by month from 300€ month salary...

Miro

[pandy]

Not cheap but ideal - my task is only one - save $80 month by month from 300€ month salary...

Having this on mind i would say then i would be more caution what i buying and how to go for other solutions - it is not clear to me why You use HDMI to everything, also it is not clear or your intention is to display or capture or display and capture at the same time.
Side to this as we talk about C128 (and it use similar video modes to classic SD) i would say there is plenty solutions.
But at some point i can only say - "not my circus and not my monkeys"

[miro@hotmail]

My answer is simple: FAST, HIGH RESOLUTION, TRUE CAPTURE
PEXHDCAP is abbreviation of Startech PCIexpress HD Capture device - this one recommended me one user of this forum (sadly, I lost old account and know not his name) - I'm very thankful! This device allows me capture HDMI up to 1080i, VGA/DVI up to 1280×1024 and RGB up to 1080i (single vertical limit 1080)
HDTV is for me great to capture, Videorecorder is SDTV, so also good
Real reason is Commodore 128 - I know that resolution 320×400 by VIC, 800×600 by VDC is lower, but no one USB grabber and my TV card too can handle interlaced modes with all of software recommended by this forum users...
PEXHDCAP works with all resolutions, modes without problems.
VDC output of C128 is problematic CGA compatible and my CGA2VGA converter displays all, but picture is shuddering...
I own also CGA2SVHS converter - no one USB grabber (my TV card too) works with interlaced modes with this converter interlaced modes, but if I connect it to my HP LCD to SVHS input - no problem - and simple conversion to RGB allows to connect it to PEXHDCAP which is always working with all modes...

Maybe you're asking why I do so many for computer of 80s of 20th century - it's my topmost hobby, I'm active user and programmer - in this case have all also unbelievable accelerator from 2MHz 8-bit CPU to 20MHz 16-bit CPU...
And why to use HD also for C128? I'm working on display device by FPGA with true color up to 1920×1200 resolution to succeedd 80's device to device of 21st century... lot of work before me - you think if I'm not crazy, but C128 I know as is possible - I'm not thinking about top ten or so - in my eyes I know so lot how I need and want and current computers are too complicated to understand them so like I know C128 - crazy hobby - satisfied man...

I know that many of you reading these lines thinking that I'm thick-headed, it's opinion of so thinking men or women.
But what I want really to write, without you all boys and girls - this forum users, I could not to know anything about all - so thank you everybody very much and I'm glad member of this forum!

Miro

[usually_quiet]

If your old username was MIRKOSOFT, then I'm the one who recommended the Startech PEXHDCAP. https://forum.videohelp.com/threads/364859-Need-recommend-Video-Grabber-higher-resoluti...n-than-7205576

I guess you figured out that composite video works via one of the PEXHDCAP's component video connections. I read a post at avsforum.com indicating the green one worked for composite. In case somebody asks in the future, is that correct?

I have always suspected S-Video would work with the PEXHDCAP using a direct connection if someone knew which pins to connect, but I have never found any posts with the pinout for the analog video port.

[pandy]

Real reason is Commodore 128 - I know that resolution 320×400 by VIC, 800×600 by VDC is lower, but no one USB grabber and my TV card too can handle interlaced modes with all of software recommended by this forum users...
PEXHDCAP works with all resolutions, modes without problems.

Sorry but this is first misunderstanding - so called interlace are usually video modes with progressive scanning but with changing video content each field - to create video really interlaced, HW need to be capable to generate particular video sequence.
Also all those modes are based on undocumented tweaks, HW modification and HW errors which mean standard equipment to capture standard video signal seem to be not appropriate - you need to go to custom HW solution (i.e. solution where you can reprogram your HW in a particular way) - i pointed to You that Bt878 family have full datasheet and it is flexible in way how it need to be programmed (check mentioned DScaler)

VDC output of C128 is problematic CGA compatible and my CGA2VGA converter displays all, but picture is shuddering...
I own also CGA2SVHS converter - no one USB grabber (my TV card too) works with interlaced modes with this converter interlaced modes, but if I connect it to my HP LCD to SVHS input - no problem - and simple conversion to RGB allows to connect it to PEXHDCAP which is always working with all modes...

At first you can easily create own RGBI converter to RGB analog (check older schematics for CGA monitors or for example well known C1084 - you can find RGBI input and as you can see on schematics conversion to analog is few resistors to form some crude DAC - but this is how it works on those days)

Maybe you're asking why I do so many for computer of 80s of 20th century - it's my topmost hobby, I'm active user and programmer - in this case have all also unbelievable accelerator from 2MHz 8-bit CPU to 20MHz 16-bit CPU...
And why to use HD also for C128? I'm working on display device by FPGA with true color up to 1920×1200 resolution to succeedd 80's device to device of 21st century... lot of work before me - you think if I'm not crazy, but C128 I know as is possible - I'm not thinking about top ten or so - in my eyes I know so lot how I need and want and current computers are too complicated to understand them so like I know C128 - crazy hobby - satisfied man...

For me too this is hobby but i prefer to do capture and later HQ postprocessing with fancy algorithm than rely on cheap bilinear or at best bicubic resize offered by most of this HW converters.

Look at http://en.wikipedia.org/wiki/Image_scaling#Pixel_art_scaling_algorithms - this is special class of resizers not the usual video oriented.

I know that many of you reading these lines thinking that I'm thick-headed, it's opinion of so thinking men or women.
But what I want really to write, without you all boys and girls - this forum users, I could not to know anything about all - so thank you everybody very much and I'm glad member of this forum!

Oh common - i have various old computers - ZX Spectrum, C64, C+4 some Atari's (and many various Amiga's - my favorite) - you are not alone but i see that you struggle with 2 different things - lack of data (then buy a cheap scope i measure video timing to understand it) second proper HW/SW solution (and as a programmer you should have no problems to understand how to reprogram HW).
Things you trying to do are not standard - they work based on assumption that CRT was flexible technology enough to deal with various differences that allow in ancient time to display video with limited number of resources.

[miro@hotmail]

Yes, my old name was MIRKOSOFT, so Thank you veeeeeeery much Usually Quiet! Works much better like I expected!
Connection Composite to Green of Component I never tried, I can, but really I need S-Video - you wrote that pinout is unknown...

Pandy, DScaler I tried, it works not too.
I can contact friend with help to create CGA (really RGBI) converter to RGB Component, I'll try, later I'll report.
Thank you for understanding my hobby...
I'm working on this device not only for me, every body can buy NEXYS4 FPGA and correct connecting is easy, problem is to programming emulation real chips.
My decision is use serial line which is available of most Commodore 8-bit computers and combine it for fastest work to have device not for C128/64 only by Expansion Port, at least in text mode for TED series etc. But future shows what is posible and what not.
CRT solution is today not good, I own no one CRT monitor for C128, for displaying I have multiinput LCDs and converters...

Thank you boys!

Miro

[pandy]

https://sites.google.com/site/h2obsession/CBM/C128/rgbi-s-video

check this site, doing small search for your problem and i've found original c1084 schematics with RGBI to RGB converter board (use 74ls368 and 74ls05) or above link (good site with various examples how easily convert video with few diodes and resistors) - i mean i see no point to spent 80$ for simple circuits that can be used with things you have already but... but with all those video modes it is difficult.
DScaler has many special video modes but it need to be used with correct HW capture chip (Bt878 or compatible) and finally you can always reprogram Bt878 to do things exactly as you wish (i assume it will be easier under linux than under windows) - some old P3 - P4 system with PCI and cheap bt878 should be ok - capture video with lossless compression and move to better machine for postprocessing), to display such video first catch datasheet to LCD controller and later reprogram video parameters to deal with your source - there is no other way - our Chinese friends don't bother with programming all possible video modes - they selecting few standardized (VESA etc) and this works - else "Out of Sync" message is in screen.

[miro@hotmail]

I know this site and also author... it's my friend and I want to ask him about RGBI2RGB - I wrote it above..
There is also circuit which was made for me - RGBI2VGA - Robert it created specially for my CGA2VGA converter 'cause original cable had not connected Instensity...

Miro

[Brad]

lossless S-Video to CVBS (composite) converter is capacitor approx 330 - 470pF - add C signal trough this capacitor to Y signal (it is lossless as it can be lossless).

Can any analog process be described as "lossless"? It seems to me even a minor reduction in S/N counts as a loss. At any rate, as soon as you combine Y+C you've altered the signal and can never recover the original precisely, therefore not lossless.

[miro@hotmail]

I meant lossless quality between Composite and S-Video

[pandy]

lossless S-Video to CVBS (composite) converter is capacitor approx 330 - 470pF - add C signal trough this capacitor to Y signal (it is lossless as it can be lossless).

Can any analog process be described as "lossless"? It seems to me even a minor reduction in S/N counts as a loss. At any rate, as soon as you combine Y+C you've altered the signal and can never recover the original precisely, therefore not lossless.

Nope - it is perfectly lossless as it doesn't remove any information - combining C and Y is lossless per se , S/N should be same (or even can be slightly better - search for information's why in electronics sometimes paralleling is used not only to increase power or current capabilities but also to reduce noise - one of many examples http://www.head-fi.org/t/541771/multiple-dac-chip-dac - paralleling is common practice when ultra low noise analog electronics is considered - cost are less relevant).

Separation of composite signal - well this not question of this topic (but it can be performed almost perfectly - http://www.jim-easterbrook.me.uk/pal/ ).
So i assume that under OP question this is perfectly lossless i.e. how to convert Y and C to CVBS.

I meant lossless quality between Composite and S-Video

This is different story but still doable especially if you can do coherent sampling and use correct algorithm.

I know this site and also author... it's my friend and I want to ask him about RGBI2RGB - I wrote it above..
There is also circuit which was made for me - RGBI2VGA - Robert it created specially for my CGA2VGA converter 'cause original cable had not connected Instensity...

So as you see there is always alternative approach - side to this quality from HW converters is usually low due of many factors - this was my point that you need to decide what is good for you - at some point you must accept limitations of HW/SW solution and used to it or design own.

[Brad]

Nope - it is perfectly lossless as it doesn't remove any information - combining C and Y is lossless per se

I have jars of blue sand and red sand. You mix them together for me and tell me nothing has been lost since it's all in there. I want my red sand back...

Separation of composite signal - well this not question of this topic (but it can be performed almost perfectly - http://www.jim-easterbrook.me.uk/pal/ ).

1. Since composite can't be displayed without separation, obviously it is inherent to the question.
2. Great, so where can we buy the BBC's solution?
3. "Almost" vs lossless.

[pandy]

I have jars of blue sand and red sand. You mix them together for me and tell me nothing has been lost since it's all in there. I want my red sand back...

Bad example - blue and red can be easily distinguished and separated - fact that sand is relatively small grain move pressure on mechanical solution, second i hope you understand concept of: phase, freqeuncy and level.

1. Since composite can't be displayed without separation, obviously it is inherent to the question.
2. Great, so where can we buy the BBC's solution?
3. "Almost" vs lossless.

Nope, OP ask about way to combine Y and C - don't blame messenger.
BBC solution can be bought from specialized companies that offer custom devices made on customer request - BBC concept is not new - if you check equation for chrominance burst you will realize that inventors of PAL introduced frequency offset to make separation possible - from mathematical perspective this is perfectly possible especially if other techniques are involved - with modern HW (usual PC processing capability) you can do this in software - so i would say any device that capture CVBS signal at least 4 * Fsc (chrominance subcarrier i.e. 4,43 in case of PAL) grayscale where chroma burst is captured to is sufficient to do remain part of processing purely in software. If some other techniques can be added to this perhaps it is possible to more than simple PAL decoding (AFAIR similar concept was used in Pixel Plus technology in Philips product).

btw CVBS can be displayed without separation http://www.techmind.org/vd/paldec.html look at the grayscale picture - it have everything to carry information about luma and chroma on grayscale.

Almost vs lossless is limited by physics and accuracy of devices. Same question related to accuracy of computation - if we use (it is sufficient for our implementation) 32 bit float why other use 128 or 256 bit float? What was first - egg or hen?

[Brad]

BBC solution can be bought from specialized companies that offer custom devices made on customer request

In other words, not something any of us can afford and certainly not for the OP's budget.

so i would say any device that capture CVBS signal at least 4 * Fsc (chrominance subcarrier i.e. 4,43 in case of PAL) grayscale where chroma burst is captured to is sufficient to do remain part of processing purely in software.

Herein lies the problem: you're happy to offer all these theoretical solutions, but you've never indicated that you've successfully implemented them.

btw CVBS can be displayed without separation http://www.techmind.org/vd/paldec.html look at the grayscale picture - it have everything to carry information about luma and chroma on grayscale.

And even that link shows that an extremely talented individual failed to fully separate Y&C.

[pandy]

In other words, not something any of us can afford and certainly not for the OP's budget.

In other words - fact that you buying DAC doesn't mean automatically that you need to buy ADC but i fully understand why you wrote something like above. Encoder is different than Decoder etc.

Herein lies the problem: you're happy to offer all these theoretical solutions, but you've never indicated that you've successfully implemented them.

Side to this - i'm not offering ready customer product as OP clearly stated that he is interesting in lossless combining Y & C - simplest way is simple add both signals - that's all.
Implementations for software decoding are well known (mostly not on PC world) - at least few commercial devices use this method - mostly frame grabbers.

And even that link shows that an extremely talented individual failed to fully separate Y&C.

This talented person use different approach (tuned bandpass filter) and don't exploit fact that chrominance spectrum is shifted from luminance spectrum with particular offset - i provided link just to show that statement "can't display CVBS without chroma and luma separation" is not accurate.

Once again - don't blame me that OP wrote how to convert separate Y & C to CVBS in lossless way.

[Brad]

i provided link just to show that statement "can't display CVBS without chroma and luma separation" is not accurate.

You know I meant in color. You may as well disprove my statement by linking me a grid displaying 1s and 0s that represent a digitally-encoded CVBS signal.

[pandy]

i provided link just to show that statement "can't display CVBS without chroma and luma separation" is not accurate.

You know I meant in color. You may as well disprove my statement by linking me a grid displaying 1s and 0s that represent a digitally-encoded CVBS signal.

Why we have this discussion? Already provided answers for OP questions - don't blame me.