Is it possible to capture with 2 inputs? And get a better signal?

[SyncroScales]

Is it possible to capture with 2 inputs? And get a better signal? Are there any capture devices that support this?

When I used my tv 'Sharp Aquos LC 26D43U' and my Sony Video Camera Recorder Hi 8: Model: CCD - TRV58, I accidentally connected the S-Video cable and the component video/mono audio cable. I watched a Hi8 tape I recorded. I remember saying this does not look too bad on my new TV. When the tape was done, I noticed that I made a mistake and connected both. After staring at the TV and plugging in each input or both and unplugging one at a time I noticed the picture does look better.

Can anyone else can explain this or try it out on a few devices? I also noticed this with the old square TV's, old video game consoles, older vcr's, etc.

When I tried to connect the TV In/ANT Out/coaxial cable it decreases the composite video resolution and very noticeably with the TV In/ANT Out/coaxial cable and S-Video.

[usually_quiet]

Is it possible to capture with 2 inputs? And get a better signal? Are there any capture devices that support this?

No, no, and no.

When I used my tv 'Sharp Aquos LC 26D43U' and my Sony Video Camera Recorder Hi 8: Model: CCD - TRV58, I accidentally connected the S-Video cable and the component video/mono audio cable. I watched a Hi8 tape I recorded. I remember saying this does not look too bad on my new TV. When the tape was done, I noticed that I made a mistake and connected both. After staring at the TV and plugging in each input or both and unplugging one at a time I noticed the picture does look better.

Can anyone else can explain this or try it out on a few devices? I also noticed this with the old square TV's, old video game consoles, older vcr's, etc.

When I tried to connect the TV In/ANT Out/coaxial cable it decreases the composite video resolution and very noticeably with the TV In/ANT Out/coaxial cable and S-Video.

Having one or more additional connections in use at the same time should not change the picture quality under normal circumstances. If the picture quality worsens when a second connection is in use, then some sort of abnormal electrical interference, for example a ground loop, is responsible.

[SyncroScales]

Thanks for the reply usually_quiet.

I am not trying to be rude. I really dont think I am seeing things. Since the signals shouldnt affect each other, what is the reason, mechanics or programming behind why this idea wont work?

[smrpix]

The only thing that might make sense is if the TV is using the composite signal as a sync reference, which some pro decks and monitors can do -- but that's a long shot. How is your "combined" picture better? Sharper, more stable, more colorful?

[usually_quiet]

I am not trying to be rude. I really dont think I am seeing things.

If the picture from one source looks better or worse when another cable is connected at the same time, it isn't normal. Something is wrong. Most likely a ground loop, RF interference, cross-talk, bad design, or a failing/defective component is responsible for the differences you are seeing. I have a TV for which every connection is used. I can turn on my cable box, both DVD recorders, the OTA antenna, and the HTPC in any combination, and the picture from any of the various source devices is unaffected by the others. I can physically disconnect something, and again, there is no difference in the picture from the remaining connection(s).

Since the signals shouldnt affect each other, what is the reason, mechanics or programming behind why this idea wont work?

Mechanics and programming prevent a capture device from using two analog signals at one time. Capture devices have only one analog to digital converter and one encoder (or no encoder if the device uses software to encode), and the analog to digital converter and encoder can only process the signal from one video connection at a time. DirectShow (Windows programming) also only allows one analog video connection to be selected at one time.

[SyncroScales]

Thanks for the info usually_quiet. Can Mac's or Linux that are not DirectShow allow more than "one analog video connection to be selected at one time"?

smrpix: The video looks more detailed. The colour is more vivid. Less pixelated or blocky.

[usually_quiet]

Thanks for the info usually_quiet. Can Mac's or Linux that are not DirectShow allow more than "one analog video connection to be selected at one time"?

The mechanics would be the same, and the drivers have to work pretty much the same way regardless of the OS. You can try physically connecting the camera to the capture device's S-Video and composite connection at the same time, but the capture device will only capture from one of them at a time. The analog signals from different connections will not be combined by the capture device either by accident or on purpose somehow and would not provide a better picture if they were.

[SyncroScales]

So is it different with TV's compared to OS's and capture devices? Do TV's allow more than one analog video connection to be selected at one time?

Is the hardware able? Eg: TV's, Cameras, Video Game Consoles, etc.

Is there any type of system or hardware that can allow multiple analog connections to be selected at one time? Maybe pro-level?

[smrpix]

It's possible with a video production switcher/mixer -- just half-dissolve between 2 inputs. Example: https://new.livestream.com/studio

But what you are seeing on your TV is an anomaly. From your description, you appear to be serendipitously benefiting, for the moment, from some sort of electronic error.

[pandy]

Yes, it can be better - doubling capture inputs in theory will improve SNR by approx. 3.0103dB (by 0.5 bit) - similar principle as for oversampling.

For example capture mono audio at the same time by stereo ADC may improve SNR by 3.0103dB (SNR for captured data not for recording itself).

CCTV capture solutions usually are capable to perform simultaneous capture on more than 2 - 4 channels at the same time.

Btw - signal improvement in Your case may be related to Y/C separation and low pass/band stop filter usually present in CVBS signal path.

[usually_quiet]

@SyncroScales I should warn you that whatever pandy has told you to do will probably not help you. It is on my ignore list for a reason. Its main purpose for joining here seems to be to show off by correcting small factual errors and starting arguments that it finds interesting, but its advice is nearly always wrong or impractical. Following it will almost certainly be a waste of your time.

[pandy]

@SyncroScales I should warn you that whatever pandy has told you to do will probably not help you. It is on my ignore list for a reason. Its main purpose for joining here seems to be to show off by correcting small factual errors and starting arguments that it finds interesting, but its advice is nearly always wrong or impractical. Following it will almost certainly be a waste of your time.

This is not beauty contest, don't behave like teenage, spoiled girl or perhaps worse - abandoned mistress...

[Cornucopia]

@pandy, your facts are just plain wrong anyway.

Take the example of capping a mono audio via 2 channels:

Ch1 has - Signal A + Noise B
Ch2 has - Signal C + Noise D

Testing for use cases:

1. If Signal A AND Signal C are identical, and Noise B and Noise D are identical, you have A=C & B=D, so A + B + C + D averaged (mixed) = (A + B + A + B) /2 = (2A + 2B) / 2= 2x (A + B) / 2 = A + B. No different from just capturing 1 channel. This example would only be possible when all those identicals are fixed, and that only happens with signals that are already. digitized. This does occur with summations in digital mixing, but otherwise is unseen.
2. If A and C are identical, but B and D are not, you would have A=C, so (A + B + C + D) / 2 = (A + B + A + D) / 2 = (A + A) + (B + D) / 2 = 2xA /2 + 1/2B + 1/2D or A + 1/2B + 1/2D. The TOTAL amount of noise is still the same proportion to the signal, it is just spread between the 2 items. This May or May Not be desirable - and this is the best case possible. If both B & D are spread-spectrum random noise anyway, they would ACT as if they are identical and be = 2xB /2 = B and no different than case 1.
3. If A and C are not identical, and B and D are not, even if you are lowering their relative weights, you also are facing the problem of modifying your A & C. Which one is correct? - they both can't be if they aren't identical. At the end of this, regardless of WHICH one is correct, you are ending up with NONE of them being fully correct. This is a scenario where you are making things worse than (possibly) just choosing one fork.
4. If A & C are not identical, but B and D are, it is even a worse scenario and you end up making the signal quieter (as well as modified) and the noise louder. This is the worst scenario case.

VHS recapping + averaging using median is an improvement on use case #2, in that it removes outliers in the mixing stage. It is this which does the "noise reduction". However, if signals A and C also already contain noise (due to analog recording stage, for example), this is identical and is NOT removed - only noise generated during the playback->capture portion of the chain can be modified.
In the case of VHS, this can amount to quite a bit. In the case of mono files, it is very, very little.

I think usually_quiet hit the nail on the head. You seem to love argument for argument's sake and take pleasure in finding exceptions in the minutiae from some theoretically perfect standpoint, disregarding the impracticality of some/much of your conjectures. This rarely proves helpful to the OP's needs.

Scott

[usually_quiet]

Unfortunate that now this will become another thread all about pandy having fun attempting to prove its theory or theories.

[pandy]

@pandy, your facts are just plain wrong anyway.

Take the example of capping a mono audio via 2 channels:

Ch1 has - Signal A + Noise B
Ch2 has - Signal C + Noise D

Testing for use cases:

1. If Signal A AND Signal C are identical, and Noise B and Noise D are identical, you have A=C & B=D, so A + B + C + D averaged (mixed) = (A + B + A + B) /2 = (2A + 2B) / 2= 2x (A + B) / 2 = A + B. No different from just capturing 1 channel. This example would only be possible when all those identicals are fixed, and that only happens with signals that are already. digitized. This does occur with summations in digital mixing, but otherwise is unseen.
2. If A and C are identical, but B and D are not, you would have A=C, so (A + B + C + D) / 2 = (A + B + A + D) / 2 = (A + A) + (B + D) / 2 = 2xA /2 + 1/2B + 1/2D or A + 1/2B + 1/2D. The TOTAL amount of noise is still the same proportion to the signal, it is just spread between the 2 items. This May or May Not be desirable - and this is the best case possible. If both B & D are spread-spectrum random noise anyway, they would ACT as if they are identical and be = 2xB /2 = B and no different than case 1.
3. If A and C are not identical, and B and D are not, even if you are lowering their relative weights, you also are facing the problem of modifying your A & C. Which one is correct? - they both can't be if they aren't identical. At the end of this, regardless of WHICH one is correct, you are ending up with NONE of them being fully correct. This is a scenario where you are making things worse than (possibly) just choosing one fork.
4. If A & C are not identical, but B and D are, it is even aworse scenario and you end up making the signal quieter (as well as modified) and the noise louder. This is the worst scenario case.

VHS recapping + averaging using median is an improvement on use case #2, in that it removes outliers in the mixing stage. It is this which does the "noise reduction". However, if signals A and C also already contain noise (due to analog recording stage, for example), this is identical and is NOT removed - only noise generated during the playback->capture portion of the chain can be modified.
In the case of VHS, this can amount to quite a bit. In the case of mono files, it is very, very little.

I think usually_quiet hit the nail on the head. You seem to love argument for argument's sake and take pleasure in finding exceptions in the minutiae from some theoretically perfect standpoint, disregarding the impracticality of some/much of your conjectures. This rarely proves helpful to the OP's needs.

Scott

All this long poem only to completely forget about fact that ADC is not transparent to signal and quantization introduce unavoidable noise - i pointed that SNR is improved but not signal SNR (as noise is part of signal but noise introduce in quantization not) - so next time Scott please involve in Your analyze for example ENOB.

Guys before trying to nail someone just read what i've wrote - sadly i would say i'm disappointed with this kind of behavior (yeah, i have this guy on ignore list but always read what he wrote and i always comment what he wrote even if i don't understand and i consider this information impractical) - perhaps for YOU this is not practical but for someone else this can be practical - world doesn't spin around your navel.

Oversampling improve SNR, paralleling converters (ADC or DAC) improve SNR but it will not improve SIGNAL SNR only CONVERSION SNR.

Side to this i provided explanation why OP may see improvement in picture quality (lack of C bandstop/notch filter in Y path).

Additionally - perhaps OP may be not able to use my information practically but it doesn't mean that He not deserve to receive information - i pointed this many times - i will not support limiting information to practical only (money making is not most important thing in world and i'm not born in USA and i can provide information not only practical from money making perspective) or to censor some "inappropriate" knowledge i.e. how to make C4.

btw large part of this "impractical" knowledge was invented in USA and it was used practically in various projects also by USA.

FYI I use practically mono signal acquisition with simple stereo PC card input to detect real life signal from HDD (acoustical) and this allow me to analyze mechanical status for HDD in device without opening device - as 3dB gain is significant.

At the end allow me to quote one of greatest people Benjamin Franklin (FYI He was born in USA also - just for case if this was forgotten).

"Freedom of speech is a principal pillar of a free government; when this support is taken away, the constitution of a free society is dissolved, and tyranny is erected on its ruins. Republics and limited monarchies derive their strength and vigor from a popular examination into the action of the magistrates. "

So don't judge what is practical or not - what can be wrote or not, especially if you judging only accordingly to your own point of view and even if you judge don't forbid others to express their opinion - i believe that anyone on this forum has equal right to express opinion and accordingly to my knowledge i didn't wrote anything against rules but even if You Scott or usually_quiet believe that i've broke rules please report this to moderator and allow him to judge.

[usually_quiet]

So is it different with TV's compared to OS's and capture devices? Do TV's allow more than one analog video connection to be selected at one time?

Is the hardware able? Eg: TV's, Cameras, Video Game Consoles, etc.

Is there any type of system or hardware that can allow multiple analog connections to be selected at one time? Maybe pro-level?

Have you considered that the benefit you are seeing might be caused by something going on in the camera's circuitry which causes a stronger composite signal to be produced when two inputs are connected, rather than anything happening at the TV end?

Actually I have a capture device with two complete sets of analog inputs that can be used at the same time, but it functions as two separate devices, although both are on one circuit board and share one PCI-e interface. It has two analog to digital converters and two encoders, one for each set of analog connections. DirectShow treats it like two capture devices. It can capture from two video inputs at one time, but from only one video connection in each set at a time. It can't combine the two video signals. Instead it outputs two separate MPEG-2 transport streams.

[Cornucopia]

That seems the most likely scenario.

Scott

[SyncroScales]

I think that theoretically composite and s-video are able to display a certain amount of signal/resolution/etc. But typically with only one connection, they display a little lower amount of the signal/resolution/etc.

What might be happening and what is included in my question is that, like you said usually_quiet, the signal is getting a boost because the tv and camcorder can handle it. At this time the composite and s-video connected simultaneously fulfill the theoretical maximum of composite and s-video. I think is it S-video, although usually_quiet has said composite.

But Microsoft's Windows DirectShow and other computer programs only allow one input to be played or recorded at a time. I don't understand why this happens and the potential of two connections boosting the signals not being taken advantage of.

I always want to boost my captures or recordings and get the best quality. So the best way to capture the recording is to use S-Video at this point.

[usually_quiet]

I think that theoretically composite and s-video are able to display a certain amount of signal/resolution/etc. But typically with only one connection, they display a little lower amount of the signal/resolution/etc.

What might be happening and what is included in my question is that, like you said usually_quiet, the signal is getting a boost because the tv and camcorder can handle it. At this time the composite and s-video connected simultaneously fulfill the theoretical maximum of composite and s-video. I think is it S-video, although usually_quiet has said composite.

But Microsoft's Windows DirectShow and other computer programs only allow one input to be played or recorded at a time. I don't understand why this happens and the potential of two connections boosting the signals not being taken advantage of.

I always want to boost my captures or recordings and get the best quality. So the best way to capture the recording is to use S-Video at this point.

You still don't get it. I think your camera has a defect of some kind or a design flaw that causes the video signal from whichever connection it is that you are having a problem with to be very weak and noisy, unless both its composite and S-Video connection are in use. The signal should be strong enough to provide a decent picture when only one video connection is used at a time. ... and it isn't possible to increase the resolution by combining both SD signals from your camera.

TVs normally use only one connection at a time, just like a capture device. You should have noticed that is is not possible to select and watch video from more that one connection at a time on a TV, unless the TV has a picture-in-picture feature, in which case you might watch video from two sources at a time, but not more, and you could choose to watch only one source.

You appear to have a theory stuck in you head and won't let go of it. After this, if you aren't convinced, I give up on trying to persuade you otherwise.

[Cornucopia]

Yes, the ONLY devices that will, can, & should accept multiple simultaneous inputs are Video Mixers. Those have to be specially designed to handle the mixing math, and they usually REQUIRE inputs that are completely synced (down to the pixel) and have identical format features, otherwise weird byproduct artifacts happen.

Scott

[SyncroScales]

Have you considered that the benefit you are seeing might be caused by something going on in the camera's circuitry which causes a stronger composite signal to be produced when two inputs are connected, rather than anything happening at the TV end?

Why would the camera show a stronger signal when two inputs are connected? When a cord is plugged into the S-video or composite and not the TV while the other one is: Nothing happens. It takes both. I don't understand what is happening. Unless it is just "the camera's circuitry".

I understand that when I plug in a composite and S-video connection a TV will only show one or the other. Is there a reason or method to explain which one is picked? (Usually the first one, sometimes depends on the device.)

[usually_quiet]

Have you considered that the benefit you are seeing might be caused by something going on in the camera's circuitry which causes a stronger composite signal to be produced when two inputs are connected, rather than anything happening at the TV end?

Why would the camera show a stronger signal when two inputs are connected? When a cord is plugged into the S-video or composite and not the TV while the other one is: Nothing happens. It takes both. I don't understand what is happening. Unless it is just "the camera's circuitry".

Based on one of your other threads, your camera apparently uses a cheap composite to S-Video converter for S-Video. The converter produces a poor quality S-Video signal from a composite signal. Possibly there is more amplification applied or a stronger composite signal is sent to the converter when both connections are used. It makes the S-Video signal stronger, although it is still a bad version of S-Video with dot-crawl artifacts because of the cheap converter. Find an electronics expert if you have to have a better, more technical explanation, but I doubt that you would understand it.

I understand that when I plug in a composite and S-video connection a TV will only show one or the other. Is there a reason or method to explain which one is picked? (Usually the first one, sometimes depends on the device.)

You originally used a remote to choose which input to watch. The TV has internal switches that control the input to be displayed based on your choice. The same input that was used the last time merely stays selected until you change it or the TV is reset.