Question about TBC in TRV-65 Hi8 Camcorder

[lollo]

Your answers are mostly BS as you don't provide anything to support your arguments.

Really?

You wrote "GV-USB2 drops/insert frames" and "GV-USB2" creates image distortion".
You posted no evidences.
On the other hand there are plenty of GV-USB2 capture available (including mine on my YouTube channel) showing the oppostite.

You wrote "Generally I see 2 options...".
You posted no evidences.
On the other hand I linked two real comparisons on the field, and mentioned that there are many other comparisons available.

So, who is bullshitting and doesn't provide anything to support his arguments?

Unlike you, I did all the tests with different cards and devices.

Really?

And who told you that I did not test the devices?

Also, it looks like you confuse component and composite signals.

Really?

You're the one who's confused. I did not mention composite. VHS/S-VHS is recorded (basically) as Y/C on tape. It should be captured as such. Few devices used as pass-through for stabilization feed the internal TBC with component-generated signal (I am not writing composite) and once processed generate the component (I am not writing composite) and the Y/C outputs from there. Only in this rare case is better to capture component (I am not writing composite) rather than Y/C.

Search old discussion between knowledgeable user "oln" and my self about it, here and on digitalFAQ for more details.

And that post was an answer to another user. I didn't ask your opinion about my choices, so please keep it for yourself.

Mine was not an opinion but a fact. I do not care what hardware/software you use, nor I wish to give a judgement; I am commenting to provide readers with the correct information.

[Sharc]

...However, you have to take into account how exactly this mechanism works.
To properly handle this signal you have to have a reference point. And this point is the bottom of the sync pulse.
So, every sync pulse gives a new reference point. It means that the between two sync pulses the referential voltage cannot and should not be changed by either side.

Correct.

However, because the capacitor is too small and the luma changes rapidly, the base line drifts within one field before reaching next sync pulse, which would set up the new proper reference point.

Here you seem to be mistaken. In my schematics and partlist of the ES-15 service manual this capacitor has a value of 100uF 1) rather than your 1uF.
Calculating the step response from sync tip (-300mV nominal) to reference white (700mV nominal for PAL) shows that the level shift during a scanline (means from one sync tip to the next sync tip (=64us)) is less than 0.4% for a nominal 75 Ohms load. In terms of 8 bit video this decay corresponds to approx. 1 step down, means luma 235->234 (worst case) for reference white. For the active picture scanline duration (52us) this shift is even less. I doubt very much that anyone would notice anything.
Panasonic engineers probably knew what they were doing
The issue may however be aggravated when a poor capture device has its own - too small - DC blocking capacitor at its input.

Anyway, to be on the safe side I agree to use the S-Video out of the SCART socket - if available. But be aware that some crap SCART->SVideo adaptor plugs are internally wildly wired, unscreened. Means they produce luma<->chroma crosstalk artifacts as I have shown in earlier posts. Choose your poison

Footnote 1): In 2 other service manuals of ES-10 and ES-15 this capacitor value was even 470uF(!). For this value the level shift between 2 successive sync tips becomes <0.1%, means less than a single step in 8bit luma -> irrelevant.

[CaptureCraft]

...However, you have to take into account how exactly this mechanism works.
To properly handle this signal you have to have a reference point. And this point is the bottom of the sync pulse.
So, every sync pulse gives a new reference point. It means that the between two sync pulses the referential voltage cannot and should not be changed by either side.

Correct.

However, because the capacitor is too small and the luma changes rapidly, the base line drifts within one field before reaching next sync pulse, which would set up the new proper reference point.

Here you seem to be mistaken. In my schematics and partlist of the ES-15 service manual this capacitor has a value of 100uF 1) rather than your 1uF.
Calculating the step response from sync tip (-300mV nominal) to reference white (700mV nominal for PAL) shows that the level shift during a scanline (means from one sync tip to the next sync tip (=64us)) is less than 0.4% for a nominal 75 Ohms load. In terms of 8 bit video this decay corresponds to approx. 1 step down, means luma 235->234 (worst case) for reference white. For the active picture scanline duration (52us) this shift is even less. I doubt very much that anyone would notice anything.
Panasonic engineers probably knew what they were doing
The issue may however be aggravated when a poor capture device has its own - too small - DC blocking capacitor at its input.

Anyway, to be on the safe side I agree to use the S-Video out of the SCART socket - if available. But be aware that some crap SCART->SVideo adaptor plugs are internally wildly wired, unscreened. Means they produce luma<->chroma crosstalk artifacts as I have shown in earlier posts. Choose your poison

Footnote 1): In 2 other service manuals of ES-10 and ES-15 this capacitor value was even 470uF(!)

Yes, I was wrong.
For composite signal, luma in dedicated S-video, luma in dedicated component Panasonic uses SAG compensation technique with 2 capacitors per line, trying to maintain the proper base line voltage.
For SCART output the signal goes straight from video/audio processor, no capacitors whatsoever, just a 75Ohm resistor.

That's from Panasonic Service Manual for DMR-ES15E(different submodels)

And yes, for SCART <-> SVideo dongles the wires are not shielded, but they are very short. So, I find it acceptable.

[Sharc]

Yes, I was wrong.
For composite signal, luma in dedicated S-video, luma in dedicated component Panasonic uses SAG compensation technique with 2 capacitors per line, trying to maintain the proper base line voltage.

Well, SAG correction lowers the low frequency cutoff of AC coupled signals without having to increase the value of the coupling capacitor excessivly. It still will not pass DC but eventually makes the up-down fluctuation less "nervous" by emulating a large capacitor.

For SCART output the signal goes straight from video/audio processor, no capacitors whatsoever, just a 75Ohm resistor.

Yes, usually still the best option.

And yes, for SCART <-> SVideo dongles the wires are not shielded, but they are very short. So, I find it acceptable

Yes, usually one can live with it.

Edit:
This note explains the SAG compensation principle nicely:
https://file.elecfans.com/web1/M00/0E/4B/pIYBAFocUdGAKLffACg1_gZ7dwA551.pdf

[AquaCCD]

The GV-USB2 is a mediocre capture card at best.....

So what other capture card(s) which are currently under active production at comparable cost are your recommendations then?

Use European PAL version (it works with NTSC too), it has a SCART connector, which has separate original S-Video signal, which is not affected by dedicated amplifier used for S-Video out

The signals are basically the same (unless they are misaligned or badly calibrated by the factory), but the S-video luma+sync (Y-wire) on the DIN/Hosiden output is AC-coupled while on the SCART output it is DC restored (fix clamped black level). This makes a difference for the subsequent device (capture card's own DC restoration and AGC). Verified with my PAL ES-15.

As far as I know nobody produces good consumer capture cards nowadays as not that many people buy them.
What market is flooded with is multiple reincarnations and variations of Ezcap, which are garbage.
Generally I see 2 options.
1. To purchase one old but tested from LordSmurf. I don't know his prices, but what I do know, they require old OS, like Windows 7, or maybe even XP.
2. To purchase semi-professional card. They are supported, they work under modern OS. They are good quality. However, they don't like "drunk" signal. You have to make sure you have good frame TBC.

I don't want to recommend anything to anyone.
However, I find it would be OK just to say what I use.
Right now I have "AJA KONE LHe+ - 0" and "Magewell Pro Capture AIO".

As far as I remember the schematic for ES-15, there is an amplifier only for dedicated S-Video out, which screws things up.
There is no amplifier for S-Video signal in SCART.

I use Windows 7. I wouldn't consider it an "old" OS even though it technically is. It's functionally the same. I also use XP. I have hardware that should reliably run anything as old as 98 but I've also had MS-DOS running on some of my old laptops so that would never be the issue.

However I cannot afford $2,000 of equipment nor would it make sense to spend that money just to capture acceptable quality video in 480i at most for sharing between friends or viewing on a consumer grade TV that only has RF/composite inputs.

[CaptureCraft]

The GV-USB2 is a mediocre capture card at best.....

So what other capture card(s) which are currently under active production at comparable cost are your recommendations then?

Use European PAL version (it works with NTSC too), it has a SCART connector, which has separate original S-Video signal, which is not affected by dedicated amplifier used for S-Video out

The signals are basically the same (unless they are misaligned or badly calibrated by the factory), but the S-video luma+sync (Y-wire) on the DIN/Hosiden output is AC-coupled while on the SCART output it is DC restored (fix clamped black level). This makes a difference for the subsequent device (capture card's own DC restoration and AGC). Verified with my PAL ES-15.

As far as I know nobody produces good consumer capture cards nowadays as not that many people buy them.
What market is flooded with is multiple reincarnations and variations of Ezcap, which are garbage.
Generally I see 2 options.
1. To purchase one old but tested from LordSmurf. I don't know his prices, but what I do know, they require old OS, like Windows 7, or maybe even XP.
2. To purchase semi-professional card. They are supported, they work under modern OS. They are good quality. However, they don't like "drunk" signal. You have to make sure you have good frame TBC.

I don't want to recommend anything to anyone.
However, I find it would be OK just to say what I use.
Right now I have "AJA KONE LHe+ - 0" and "Magewell Pro Capture AIO".

As far as I remember the schematic for ES-15, there is an amplifier only for dedicated S-Video out, which screws things up.
There is no amplifier for S-Video signal in SCART.

I use Windows 7. I wouldn't consider it an "old" OS even though it technically is. It's functionally the same. I also use XP. I have hardware that should reliably run anything as old as 98 but I've also had MS-DOS running on some of my old laptops so that would never be the issue.

However I cannot afford $2,000 of equipment nor would it make sense to spend that money just to capture acceptable quality video in 480i at most for sharing between friends or viewing on a consumer grade TV that only has RF/composite inputs.

Sorry, where this $2,000 figure is coming from?

[AquaCCD]

Not sure where this USD 1500 came from.
Quick search on eBay shows 10 units available at US $44.00.

I assume the other poster did what I did and just pasted "AJA KONE LHe+ - 0" into a search engine. The first result for me was a page on B&H Photo's website listing a device for over $1300.

I did look it up on eBay just now, and the first results are also for over $1300. I will admit where it shows you "results matching fewer words" there are some AJA brand PCI cards that are around $44. Could you be more specific about which model of card you're using or clarify this for us, because when we just look up the string you posted it turns up a lot of very expensive results and I'd guess most of us aren't familiar enough with this hardware to be able to easily recognize the difference between model names given that they're numbers and digits.

A $50 PCI card is definitely doable for me and I would consider it alongside a USB capture device. Only problem for me right now is my desktop computer is broken, and while I do plan to fix it sometime in the future, getting it ready for video capture soon would require me to spend more money on the other things I need for it to work again.

[AquaCCD]

The GV-USB2 is a mediocre capture card at best.....

So what other capture card(s) which are currently under active production at comparable cost are your recommendations then?

Use European PAL version (it works with NTSC too), it has a SCART connector, which has separate original S-Video signal, which is not affected by dedicated amplifier used for S-Video out

The signals are basically the same (unless they are misaligned or badly calibrated by the factory), but the S-video luma+sync (Y-wire) on the DIN/Hosiden output is AC-coupled while on the SCART output it is DC restored (fix clamped black level). This makes a difference for the subsequent device (capture card's own DC restoration and AGC). Verified with my PAL ES-15.

As far as I know nobody produces good consumer capture cards nowadays as not that many people buy them.
What market is flooded with is multiple reincarnations and variations of Ezcap, which are garbage.
Generally I see 2 options.
1. To purchase one old but tested from LordSmurf. I don't know his prices, but what I do know, they require old OS, like Windows 7, or maybe even XP.
2. To purchase semi-professional card. They are supported, they work under modern OS. They are good quality. However, they don't like "drunk" signal. You have to make sure you have good frame TBC.

I don't want to recommend anything to anyone.
However, I find it would be OK just to say what I use.
Right now I have "AJA KONE LHe+ - 0" and "Magewell Pro Capture AIO".

As far as I remember the schematic for ES-15, there is an amplifier only for dedicated S-Video out, which screws things up.
There is no amplifier for S-Video signal in SCART.

I use Windows 7. I wouldn't consider it an "old" OS even though it technically is. It's functionally the same. I also use XP. I have hardware that should reliably run anything as old as 98 but I've also had MS-DOS running on some of my old laptops so that would never be the issue.

However I cannot afford $2,000 of equipment nor would it make sense to spend that money just to capture acceptable quality video in 480i at most for sharing between friends or viewing on a consumer grade TV that only has RF/composite inputs.

Sorry, where this $2,000 figure is coming from?

Sorry, please see my post above. Looks like there was some confusion. I'll admit I was initially a bit annoyed someone would suggest something so expensive but I think it was just a misunderstanding.

In my defense, it's very common on a lot of forums for people to recommend unreasonably expensive gear for various things so that's what I assumed was happening. But like I said in my previous post after looking at partial matches on eBay the prices seem reasonable for the task I want, decent but not archival quality.

[CaptureCraft]

Not sure where this USD 1500 came from.
Quick search on eBay shows 10 units available at US $44.00.

I assume the other poster did what I did and just pasted "AJA KONE LHe+ - 0" into a search engine. The first result for me was a page on B&H Photo's website listing a device for over $1300.

I did look it up on eBay just now, and the first results are also for over $1300. I will admit where it shows you "results matching fewer words" there are some AJA brand PCI cards that are around $44. Could you be more specific about which model of card you're using or clarify this for us, because when we just look up the string you posted it turns up a lot of very expensive results and I'd guess most of us aren't familiar enough with this hardware to be able to easily recognize the difference between model names given that they're numbers and digits.

A $50 PCI card is definitely doable for me and I would consider it alongside a USB capture device. Only problem for me right now is my desktop computer is broken, and while I do plan to fix it sometime in the future, getting it ready for video capture soon would require me to spend more money on the other things I need for it to work again.

I see.
First, I screwed up a little bit with the name. It is, of course "AJA KONA LHe+ - 0"
Second, yes the $44 card you found is the one.
However, there is catch.
The card there does not come with breakout cable - you need to find it and purchase separately.
And there is another catch - this breakout cable comes with BNC connectors for video and SLR connectors for audio. That means you would need to purchase RCA-to-BNC and RCA-to-SLR adapters. They're not expensive, but all of this just adds up.
The software is available on their website. You don't need to register there to download it, just reed the text at the bottom when message pops up.
Keep in mind this card captures full range of video, which has pros and cons.
The good thing is that it wouldn't crash you "extra" black and white.
The bad thing is that you might think you're capturing proper video.

[Alwyn]

it's very common on a lot of forums for people to recommend unreasonably expensive gear for various things so that's what I assumed was happening

I agree. The GV-USB2 is a great digitiser, it just works, with video quality proven to be the same as the 710-USB. Nor do you need a flash VCR; condition these days is more important than "quality". What you do need is a stabiliser, the most cost-effective being a DVD-recorder with those properties eg an ES-10, ES-15 or other Panasonics with DIGA and similar Sonys, and Pioneers.

And yes, depending on Capture-craft's exact cards, it would be very easy to wrack up a close-to or over $2000 bill. A "real" frame TBC will cost you north of $1000 (if it works at all) and an S-VHS VCR with a TBC will set you back at least $450; check out vcrshop.com. Add on $200 for a 710-USB from you-know-who and you're pushing some serious dollars. Cost-benefit-analysis? Fail, IMO.

[Sharc]

Keep in mind this card captures full range of video, which has pros and cons.

No cons. A good capture card should allow to exploit the full range (aka "capturable range" 1....254 in 8bit realm). Clipping the luma at 16 and 235 by the card (or any other device in your workflow) is usually a disaster - except perhaps for overshoots (short spikes) caused by ringing and halos.
Final levels adjustments to comply to "limited (aka TV) range" - as far as neccessary for distribution or standards compliance - should be done in post anyway, or within certain constraints (coarse adjustments) with the proc-amp settings of the card, depending on which device causes the clipping.
Bottom line: Avoid clipping of the real video like the plague - except perhaps for overshooting artifacts (short spikes).

The good thing is that it wouldn't crash you "extra" black and white.

Sic !

[CaptureCraft]

Keep in mind this card captures full range of video, which has pros and cons.

No cons. A good capture card should allow to exploit the full range (aka "capturable range" 1....254 in 8bit realm). Clipping the luma at 16 and 235 by the card (or any other device in your workflow) is usually a disaster - except perhaps for overshoots (short spikes) caused by ringing and halos.
Final levels adjustments to comply to "limited (aka TV) range" - as far as neccessary for distribution or standards compliance - should be done in post anyway, or within certain constraints (coarse adjustments) with the proc-amp settings of the card, depending on which device causes the clipping.
Bottom line: Avoid clipping of the real video like the plague - except perhaps for overshooting artifacts (short spikes).

The good thing is that it wouldn't crash you "extra" black and white.

Sic !

I would strongly disagree about post-digitizing adjustments of luma, like contrast.
You might won't see that visually, but mathematically such video would be a disaster.
If you take a look at a histogram, you would see equally spread spikes. The histogram would look like a comb instead of having a smooth top.
It's called combing, when all pixels with a given luma are added to the neighbour luma. Like all pixels with luma 201 will be reduced by 1 point and will be added to pixels with luma 200 in the histogram. However, original luma 200 and 201 would stay as is.
So, you'll get double amount of pixels with luma 200 and nothing with luma 201. Hence the spikes.
The only proper way to squeeze your video into the legal range is to adjust it before digitizing.
That's the one of the reasons I have Sony DVD recorder in my chain.
So, before digitizing a video I do quick scan to check for too dark and too bright scenes and adjust the settings to make sure the video stays within legal range.
Only after that I do the actual capture.

[Sharc]

Keep in mind this card captures full range of video, which has pros and cons.

No cons. A good capture card should allow to exploit the full range (aka "capturable range" 1....254 in 8bit realm). Clipping the luma at 16 and 235 by the card (or any other device in your workflow) is usually a disaster - except perhaps for overshoots (short spikes) caused by ringing and halos.
Final levels adjustments to comply to "limited (aka TV) range" - as far as neccessary for distribution or standards compliance - should be done in post anyway, or within certain constraints (coarse adjustments) with the proc-amp settings of the card, depending on which device causes the clipping.
Bottom line: Avoid clipping of the real video like the plague - except perhaps for overshooting artifacts (short spikes).

The good thing is that it wouldn't crash you "extra" black and white.

Sic !

I would strongly disagree about post-digitizing adjustments of luma, like contrast.
You might won't see that visually, but mathematically such video would be a disaster.
If you take a look at a histogram, you would see equally spread spikes. The histogram would look like a comb instead of having a smooth top.
It's called clipping, when all pixels with a given luma are added to the neighbour luma. Like all pixels with luma 201 will be reduced by 1 point and will be added to pixels with luma 200 in the histogram. However, original luma 200 and 201 would stay as is.
So, you'll get double amount of pixels with luma 200 and nothing with luma 201. Hence the spikes.
The only proper way to squeeze your video into the legal range is to adjust it before digitizing.
That's the one of the reasons I have Sony DVD recorder in my chain.
So, before digitizing a video I do quick scan to check for too dark and too bright scenes and adjust the settings to make sure the video stays within legal range.
Only after that I do the actual capture.

Be assured I know this mechanism. The spikes and gaps in the histogram are due to integer rounding when doing gain adjustments after digitization with finite bitdepth. Not nice, but one would hardly notice it for noisy VHS. It has has however nothing to do with signal clipping and is usually pretty harmless compared to clipping.
And even when you adjust the video in analog domain before digitization to stay within the legal range (luma and chroma) this does not warrant that you won't have clipped RGBs after converting your legal YUV to RGB for viewing on a monitor or TV. Keywords: The RGB block in the YUV cube / legal YUV producing invalid RGB. Again, you may not notice it as the color distortions are usually subtle, especially when only one of the R,G or B component gets clipped. Plenty of tutorials and posts about it.

[CaptureCraft]

Keep in mind this card captures full range of video, which has pros and cons.

No cons. A good capture card should allow to exploit the full range (aka "capturable range" 1....254 in 8bit realm). Clipping the luma at 16 and 235 by the card (or any other device in your workflow) is usually a disaster - except perhaps for overshoots (short spikes) caused by ringing and halos.
Final levels adjustments to comply to "limited (aka TV) range" - as far as neccessary for distribution or standards compliance - should be done in post anyway, or within certain constraints (coarse adjustments) with the proc-amp settings of the card, depending on which device causes the clipping.
Bottom line: Avoid clipping of the real video like the plague - except perhaps for overshooting artifacts (short spikes).

The good thing is that it wouldn't crash you "extra" black and white.

Sic !

I would strongly disagree about post-digitizing adjustments of luma, like contrast.
You might won't see that visually, but mathematically such video would be a disaster.
If you take a look at a histogram, you would see equally spread spikes. The histogram would look like a comb instead of having a smooth top.
It's called clipping, when all pixels with a given luma are added to the neighbour luma. Like all pixels with luma 201 will be reduced by 1 point and will be added to pixels with luma 200 in the histogram. However, original luma 200 and 201 would stay as is.
So, you'll get double amount of pixels with luma 200 and nothing with luma 201. Hence the spikes.
The only proper way to squeeze your video into the legal range is to adjust it before digitizing.
That's the one of the reasons I have Sony DVD recorder in my chain.
So, before digitizing a video I do quick scan to check for too dark and too bright scenes and adjust the settings to make sure the video stays within legal range.
Only after that I do the actual capture.

Be assured I know this mechanism. The spikes and gaps in the histogram are due to integer rounding when doing gain adjustments after digitization. Not nice, but one would hardly notice it for noisy VHS. It has has however nothing to do with signal clipping.
And even when you adjust the video in analog domain before digitization to stay within the legal range (luma and chroma) this does not warrant that you won't have clipped RGBs after converting your legal YUV to RGB for viewing on a monitor or TV. Keywords: The RGB block in the YUV cube / legal YUV producing invalid RGB. Plenty of tutorials and posts about it.

I used wrong term - not clipping, combing.
Anyway, I consider this as a catastrophe and I do everything to avoid it.

As for RGB <> YUV conversion, just make your range a lite bit narrower, don't push it to the tilt.

[lollo]

So, before digitizing a video I do quick scan to check for too dark and too bright scenes and adjust the settings to make sure the video stays within legal range.
Only after that I do the actual capture.

The best practice is the opposite: capture in the range allowed by the capture card reducing to minimum the modification to the histogram performed by the procamp (destructive), and do all levels adjustement in post-processing (under user control).

As Sharc said, we discussed about all this many times in the forum (he actually prefers to capture in the limited range, but the concepts are the same).

Few references:
https://forum.videohelp.com/threads/402267-Adjusting-capture-card-levels-not-sure-if-i...ht#post2624452
https://forum.videohelp.com/threads/401349-GV-USB2-Capture-Stick-What-s-In-It
https://forum.videohelp.com/threads/401639-Interpreting-VirtualDub-Capture-Histogram

[Sharc]

As Sharc said, we discussed about all this many times in the forum (he actually prefers to capture in the limited range, but the concepts are the same).

Exactly. My usual recommendation to capture in the limitied range is mainly for safeguarding newbies from running into certain pitfalls.
There is no disagreement on the principles.
https://forum.videohelp.com/threads/418254-Info-Hauppauge-USB-Live-2-capturable-range#post2780708

[CaptureCraft]

Some videos have wider than full range.
So, you guys can keep doing the post-processing, I'll keep the pre-digitizing adjustment to avoid any post-capture changes.

[lollo]

It wasn't meant to convince you; it was just to show the general reader what is right and wrong.

[CaptureCraft]

It wasn't meant to convince you; it was just to show the general reader what is right and wrong.

I do not see anything right in your method.
It looks easier, yes, but that doesn't make it right.
I would rather let general reader to decide which method has better fit for their needs.

[lollo]

I would rather let general reader to decide which method has better fit for their needs.

It wasn't meant to convince the general reader either; it was just to show what is right and wrong.

[CaptureCraft]

I would rather let general reader to decide which method has better fit for their needs.

It wasn't meant to convince the general reader either; it was just to show what is right and wrong.

Explicitly telling to someone that something is right or wrong is imposition of conviction.

[Sharc]

.... As for RGB <> YUV conversion, just make your range a lite bit narrower, don't push it to the tilt.

Unfortunately that may not help. Legalization is more complicated. Take for example YUV (225,210,180). It is perfectly within legal YUV limits, right? But it leads to heavily clipped R and B upon conversion to RGB. Your TV or monitor can't display YUV. The YUV data have to be converted ("matrixed") to RGB by the renderer. Keyword: The RGB block in the YUV cube.
Anyway, I don't want to overstress this subject/problem. Just a reminder.
Time to here.

[CaptureCraft]

.... As for RGB <> YUV conversion, just make your range a lite bit narrower, don't push it to the tilt.

Unfortunately that may not help. Legalization is more complicated. Take for example YUV (225,210,180). It is perfectly within legal YUV limits, right? But it leads to heavily clipped R and B upon conversion to RGB. Your TV or monitor can't display YUV. The YUV data have to be converted ("matrixed") to RGB by the renderer. Keyword: The RGB block in the YUV cube.
Anyway, I don't want to overstress this subject/problem. Just a reminder.
Time to here.

Yes, you can't properly convert YUV (225,210,180) to RGB.
However, I wouldn't worry about it at all as there is not a chance you can really shoot that on video.
And even if you can inject that via built-in camcorder character generator, such chroma would be distorted on the tape.
On the other hand, maybe someone will map it tomorrow to HDR, which is theoretically possible .

[lollo]

Explicitly telling to someone that something is right or wrong is imposition of conviction.

I didn't explicitly tell anyone anything; I was just writing what's right or wrong.

However, I wouldn't worry about it at all as there is not a chance you can really shoot that on video.

Wrong.

And even if you can inject that via built-in camcorder character generator, such chroma would be distorted on the tape.

Wrong (Y, and U/V combined to C in analog domain).

[Sharc]

The only proper way to squeeze your video into the legal range is to adjust it before digitizing.
That's the one of the reasons I have Sony DVD recorder in my chain.
So, before digitizing a video I do quick scan to check for too dark and too bright scenes and adjust the settings to make sure the video stays within legal range.
Only after that I do the actual capture.

That's ok. But out of curiosity: how do you adjust the analog video levels to stay withing legal range using your SONY DVD recorder? I don't have this model, but from the manual I didn't find an option to adjust analog levels in analog domain i.e. before digitizing. Maybe I am missing something .....

Edit: Oh I see, page 120ff. It lists some rudimentary options like White AGC, White enhancer, Black enhancer, Black level (sets or removes the NTSC 7.5 IRE pedestal only), Hue and Chroma level. It is also not clear to me whether these (or which) settings act in analog or digital domain. But again, I might be missing something ....

[CaptureCraft]

The only proper way to squeeze your video into the legal range is to adjust it before digitizing.
That's the one of the reasons I have Sony DVD recorder in my chain.
So, before digitizing a video I do quick scan to check for too dark and too bright scenes and adjust the settings to make sure the video stays within legal range.
Only after that I do the actual capture.

That's ok. But out of curiosity: how do you adjust the analog video levels to stay withing legal range using your SONY DVD recorder? I don't have this model, but from the manual I didn't find an option to adjust analog levels in analog domain i.e. before digitizing. Maybe I am missing something .....

Edit: Oh I see, page 120ff. It lists some rudimentary options like White AGC, White enhancer, Black enhancer, Black level (sets or removes the NTSC 7.5 IRE pedestal only), Hue and Chroma level. It is also not clear to me whether these (or which) settings act in analog or digital domain. But again, I might be missing something ....

White enhancer and Black enhancer. Everything else should be off or in neutral position.
They work like an asymmetrical contrast.
If you slide a slider to the right (positive territory), then corresponding brightness is pushed in, i.e. becomes closer to the centre of histogram.
If you slide a slider to the left (negative territory), then corresponding brightness is pulled out, i.e. becomes further away from the centre of histogram.
This way you make sure your video doesn't flow outside of the legal range.
All these settings act in digital domain (10-bit). However, there are no trace of any combing whatsoever, probably because of further converting to analog signal and/or because most likely Sony uses Digital Dithering as they have a high-end chip there (MC10050F1-505-LU1-A).
Anyway, these changes are the last resort.
Before that I changed the output level in my VCR.
I have JVC VCR, and I've created my own service remote control for it, which allowed me to make some important changes, including output level of luma.
The VCR's luma gain control has low-resolution granularity - you don't need oscilloscope to control it - it took me only two discrete decrements to bring the signal from an overmodulated state down to broadcast-safe limits (legal range).
After that I only adjust levels per record using Sony DVD.
That's pretty useful especially for dubs made from different camcorders to different VCRs at different times.

[CaptureCraft]

Explicitly telling to someone that something is right or wrong is imposition of conviction.

I didn't explicitly tell anyone anything; I was just writing what's right or wrong.

That's exactly what you just did!

However, I wouldn't worry about it at all as there is not a chance you can really shoot that on video.

Wrong.

BS

And even if you can inject that via built-in camcorder character generator, such chroma would be distorted on the tape.

Wrong (Y, and U/V combined to C in analog domain).

Another BS

See, it looks like you argue just for the sake of arguing. You don't provide any real arguments. There is either nothing or complete BS.

For example, the YUV (225,210,180) in real life would be a glowing pink neon sign with the brightness roughly from 3000 to 5000 nits.
However, CCD censor is designed to handle 100 nits (120 nits in peak), which corresponds to luma = 235.
Do you really think CCD sensor will handle 3000 nits?!?

About the recording the "impossible" YUV.
The video is recorded using amplitude modulation, it means Y and C electrical voltages physically stack on top of each other.
In this case both luma and chroma close to their max values.
Do you really think the circuit would let almost double voltage to go through and hit the heads? And that the tape can record a signal of any amplitude without a distortion?

Anyway, i hope this is the last time I respond to your messages.
I'm really tired of your BS.

[AquaCCD]

Although I think the PCI card would work for me, I travel a lot and often only have my laptop with me, so for now, I think I'm going to go with the GV-USB2, with the possibility of upgrading my desktop computer in the (hopefully near) future when I am not on the go so much. I may post an update later with some sample footage.

[dellsam34]

Sorry that your thread went of the rail, This has been happening a lot lately. Other forums don't tolerate this, we are taking it for granted here.

[lollo]

See, it looks like you argue just for the sake of arguing. You don't provide any real arguments. There is either nothing or complete BS.

Not really, I just write what is rigth or wrong.

For example, the YUV (225,210,180) in real life would be a glowing pink neon sign with the brightness roughly from 3000 to 5000 nits.
However, CCD censor is designed to handle 100 nits (120 nits in peak), which corresponds to luma = 235.
Do you really think CCD sensor will handle 3000 nits?!?

Writing "Y=235 corresponds to “100 nits” or that "YUV (225,210,180) would represent a “3000–5000‑nit neon sign" is the real BS.
Luminance (cd/m², nits) is a physical measurement of real light. These two quantities are not directly linked. Luma depends on RGB→YUV conversion, gamma, color matrix, sensor processing, etc. and depends on the actual amount of light hitting the sensor.
Moreover, CCD sensors are not limited to “100 nits = Y235”, whose saturation is determined by other parameters not by a fixed luminance threshold in nits.

3000–5000 nits corresponding to YUV (225,210,180) in only in your fantasy, because you confuse luma with luminance, assign CCD sensors arbitrary limits in nits, and interpret YUV values as physical brightness.

About the recording the "impossible" YUV.
The video is recorded using amplitude modulation, it means Y and C electrical voltages physically stack on top of each other.
In this case both luma and chroma close to their max values.
Do you really think the circuit would let almost double voltage to go through and hit the heads? And that the tape can record a signal of any amplitude without a distortion?

The luminance FM carrier and the chroma color‑under subcarrier do not “stack” their voltages. Both are amplitude‑limited, filtered, and modulated before reaching the heads.
The tape never receives a signal whose amplitude depends on a specific YUV triplet.
All legal colors, including YUV (225,210,180), are mapped into the fixed recording ranges.
Any chroma distortion on VHS or S‑VHS comes from bandwidth limitations and color‑under processing, not from excessive electrical amplitude caused by saturated colors.

Anyway, i hope this is the last time I respond to your messages.

Then stop writing non sense and bullshit.

I'm really tired of your BS.

Again, who is bullshitting?

Sorry that your thread went of the rail, This has been happening a lot lately. Other forums don't tolerate this, we are taking it for granted here.

I agree. I will stop here