This is working for me PAL VHS capture uncompressed

[Eric-jan]

wow ! do i see some mixed up things... VHS PAL was never about pixels, but about lines. different CRT TV's showed you a portion of that,(depending correct factory settings) 4:3 was the visible aspect ratio area, The DVD standard has a greater visible area, 5:4 fitting 4:3 into that by equal ratio means.
The nasty thing is that capture (output) formats are based on the DVD standard, and one cannot set the capture area like one can do with a flatbed scanner.
so pixels are wasted and not taking part of the quality of the image.
One have to choose cropping on full screen with pillar bars = resolution loss, or masking (not full screen) = smaller image & no resolution loss.
(based on 4:3 visible area)

or did i miss something ?

[DB83]

I know this reply just fouls the topic all the more but sometimes it is necessary.


The only thing that is without dispute is that the PAL analogue broadcast standard is based around 576 visible lines (from 625 actual lines)


I know I read somewhere why the dvd-video standard settled on 704/720 horizontal pixels but I forget where. Yet the vertical pixels are still based on the visible lines of the analogue system. But a dvd can be broadcast on an analogue system and those 576 lines on your tv do not suddenly become 540. They remain at 576.


But it is nonsense to even suggest that analogue capture is based around the dvd standard. There were video cards available years before dvd was developed. And I had a card that could capture at 640*480. Go figure that one.

[lordsmurf]

The nasty thing is that capture (output) formats are based on the DVD standard,

No.
DVD-Video uses the same D1 specs as everything else SD.

[Eric-jan]

The nasty thing is that capture (output) formats are based on the DVD standard,

No.
DVD-Video uses the same D1 specs as everything else SD.

Then why do you say no ? i believe DB83 has corrected me correctly, D1 also was a standard in CCTV.

[dellsam34]

Then why do you say no ? i believe DB83 has corrected me correctly, D1 also was a standard in CCTV.

He said no because it's the other way around. DVD and capture cards are based on the D1 standard.

As far as I know all capture card chipsets are based on the D1 or ITU-R 601 (orig. CCIR-601) standard aka Rec. 601 way before DVD was on the scene which later it used the D1 standard for its non widescreen format. The easiest way to find out your capture card native resolution is to look at the chipset datasheet, If it says sampling at ITU 601 than it captures at 720(704+16 balck pixels) x the color standard vertical resolution. Forget the capture software resolution options, those are just output options.

[Eric-jan]

The only thing that is without dispute is that the PAL analogue broadcast standard is based around 576 visible lines (from 625 actual lines)

Never did hear that one

[DB83]

No need to take my word for it.


https://en.wikipedia.org/wiki/576i


Article also mentions Rec. 601 referred to above.


You may not be old enough to remember the previous standard of 405 lines

[Eric-jan]

No need to take my word for it.


https://en.wikipedia.org/wiki/576i


Article also mentions Rec. 601 referred to above.


You may not be old enough to remember the previous standard of 405 lines

405 lines? wasn't that used only in the UK ? they had VHS already then ? Analog PAL has only 625 lines max. never heard of any more anount then that.

[DB83]

No idea. A European would have to research as to what their analogue broadcasts were prior to 625 lines.


And if I read correctly both 405 lines and 625 lines are themselves based around the electric power distribution of 50hz (which, ultimately, converts to fields and then fps (in the digital realm) )


The wiki does mention that 625 lines was 'available' from the 1950's. Yet it only came to the fore in the UK when BBC2 started. Tv Sets capable of receiving it were then expensive. I do recall the first set we had to receive this. Probably expensive with the components liable to fail if you dared to breathe on them (which they regularily did)

[dellsam34]

UK is just like the US they've gone through different TV standards before they settled on their final ones, 405 lines was just a TV broadcast standard even when it became possible to record it it was recorded on normal Betamax and VHS machines so it is not a tape format per se.

[dellsam34]

The only thing that is without dispute is that the PAL analogue broadcast standard is based around 576 visible lines (from 625 actual lines)

Never did hear that one

You live in a PAL country and never knew that PAL is 625 line TV standard?

[Eric-jan]

The only thing that is without dispute is that the PAL analogue broadcast standard is based around 576 visible lines (from 625 actual lines)

Never did hear that one

You live in a PAL country and never knew that PAL is 625 line TV standard?

Well... the number of visible lines that is, they did not know that in the analog time according to me, it all depended on what you saw on your CRT,
I see lots of different active area's on my VHS tapes, so i guess depending on used equipment lines where also of different length.
(I also noticed even my Sony Bravia is masking away more of the image info in 4:3 "normal" mode than i like.)
PAL is always said to be 625 lines and two fields of each 25 lines, which is always a problem to make something nice of it, and even these lines have dots, which leaves a poor resolution on a CRT.

[dellsam34]

576 vertical lines was defined by the D1 standard when they started migrating analog video contents to digital tape formats such as digibeta by Sony. 576 didn't exist in the analog video world and CRT displays.

[DB83]

^^ Are you then disputing the wiki article I referred to above ?


"In analogue 49 additional lines without image content are added to the displayed frame of 576 lines to allow time for older cathode ray tube circuits to retrace for the next frame,[2] giving 625 lines per frame."

[dellsam34]

Yes, that's totally incorrect, 625 was set in the 50's before digital and before any VCR was invented. You are aware that wikipedia can be written and edited by anyone including yourself, right? It's not a white paper.

This too came from wikipedia:

D-1 resolution is 720 (horizontal) × 486 (vertical) for NTSC systems and 720 × 576 for PAL systems; these resolutions come from Rec. 601.

So which wikipedia you believe in?

Here is a nice theory:
http://martin.hinner.info/vga/pal.html

[DB83]

My friend. I am not here to argue with you and I readily accept that wiki pages can be written by any 'crank'.


Even so, I often see some 'editorial' than entries require citations. I have not checked down the screen but there are typically external links that backup the 'facts' presented.


But if this info is incorrect then one would also have to dispute the other one that discussed 405 lines when it mentions that 377 were picture with the rest for timing.


But if it is not 576 lines for picture then what was it ? It could not have been the full 625 else how could, for example, teletext be transmitted over analogue systems. Or am I just have a 'senior' moment.

[dellsam34]

Again it wasn't 576 until the arrival of digital that put a number on it. The excessive overscan of CRT TV's barely resolved 450 vertical lines across the silver screen, so it wasn't a big deal back then how many true active video lines, Only people who have seen the full video frame are the repairman who have access to special monitors that have the ability to shift the frame half way horizontally and vertically or a crazy person like me who changed the overscan to almost zero by re-adjusting the H and V pots of my TV. But when digital came along they had to define what is the active video area and what is not to be able to capture a video frame. Hope this helps.

[DB83]

Ok. But are you not 'influenced' by the US system where there were only 525 total lines ?


https://en.wikipedia.org/wiki/480i


Or is this article also wrong ? Maybe written by the same 'crank'


The logical reasoning is that digital followed the analogue 'active' picture otherwise how did they settle on 480/576 ?. Numbers are not simply plucked out of thin air.


There really must be some definitive information out there. So I found, I think, the technical info


https://www.itu.int/net/ITU-R/index.asp?category=information&rlink=rec-601&lang=en


Led me to:


https://www.itu.int/dms_pubrec/itu-r/rec/bt/R-REC-BT.601-5-199510-S!!PDF-E.pdf


But little, to my untrained eyes, that specifically mentions 480/576 (480 samples per digital line appears common to both but that number may just be co-incidental)


But what I really need to find is some independent info on analogue broadcast systems since if I read the Rec. correctly it is more concerned with the conversion of analogue colour primaries to digital ones.

[dellsam34]

A simplified theory for both PAL and NTSC is linked in post 75. even 480/576 does include some head switch noise, The analog video area was not consistent from one broadcast to another so they decided to go with a generous number so they don't cut off the actual video, It was even more inconsistent horizontally to the point they decided to reserve 8 black pixels on each side of the frame (704 + 16) so they make sure no video is being cut off, And then later on D8, DV and DVD used the whole 720 pixels for video.

[DB83]

I just found this


https://www.causewaysecuritysolutions.com/resolution-calculated-tv-lines.html

[DB83]

And this:


http://datagenetics.com/blog/april12018/index.html


which may have effectively nailed the 576 maximum visible lines (625 - [25*2] rounded up to an even number)


which also proves NTSC at a maximum resolution of 486 (525 - [20*2] rounded up)


And the earlier Rec. specific to 625 Analog(ue)


https://www.itu.int/dms_pubrec/itu-r/rec/bt/R-REC-BT.470-6-199811-S!!PDF-E.pdf


Which I believe does mention the vertical blanking used in these formulae.

[dellsam34]

Yes the engineers back in the 80's when they proposed rec601 standards they didn't come up with 576 and 480 (486 SMPTE) magically, they were based on math formulas.

[DB83]

Well, my friend, with the last two links, and that Rec. , which pre-dates 601 by some, I am quite convinced, even if I can not follow the algebra, there is a definite correlation between the analogue standard and the digital one.


I will leave this discussion now unless someone wishes to break down the maths in to some more simple equations bur also having regard to the info in the earlier Rec.

[Nands67]

Very interesting posts and thanks for the info.

I am lost with the actual argument?

I have not found a stable program that can capture from USB device VHS PAL uncompressed video & audio interleaved with audio & video in sync.

Have been playing with FFmpeg but the media info shows 50 FPS progressive and the audio has constant click/drop out.

I probably have the wrong script settings.

Any advice other than Virtualdub, Virtual VCR?

[DB83]

Welcome back.


I think I already stated way back when that capturing VHS as uncompressed is a waste of resources. You really should consider a lossless codec such as lagarith. Your file sizes will be smaller. Visual quality will be no different. And your files will be easier to edit with no compression loss (same as with uncompressed) until you re-encode to a delivery format.

[Nands67]

Lol DB83

You have a wealth of information and I appreciate that.

But if you reflect I am not concerned with file size.
I don’t want a lossless algorithm.
I want raw data.
Then I can do whatever I want.

[Nands67]

I agree with what you are saying, I am fully aware with the lossless codecs but.
The point is what I said.

[lordsmurf]

But if you reflect I am not concerned with file size.
I don’t want a lossless algorithm.
I want raw data.
Then I can do whatever I want.

You don't get it.

"file size" isn't just about size on disk. For example, the processing time is always 2x+, and is a severe bottleneck.

There is no such thing as "raw data". It exists in some format, in some colorspace, and must be contained in some wrapper.

This isn't like RAW and JPEG, like in photo.

Lossless is like the RAW, H.264 is like the JPEG. The unadulterated sensor data is the uncompressed "raw file", and doesn't exist in most settings, as it's really not usable.

Uncompressed isn't used professionally, or by home users.

[Nands67]

Wow that’s A very negative response and who said anything about the codec you referred to H.264 as being any ware in the discussion

[Nands67]

I am aware of process time, cpu & threads etc

This should be easily be achievable for capture SD.