Line-Out vs. Headphone on portable CD

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Dave Moll, Andrew Davie, Steve Anderson

Line-Out vs. Headphone on portable CD

Postby Andrew Davie » Tue May 29, 2007 1:11 am

The attached image shows results of the same image with the source being either line-out or headphone-out on a portable CD player. I really can't see any problems with the headphone-out on this CD; I wonder if others are the same?
Attachments
lineheadphone.jpg
lineheadphone.jpg (55.4 KiB) Viewed 13751 times
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1590
Joined: Wed Jan 24, 2007 4:42 pm
Location: Queensland, Australia

Postby DrZarkov » Tue May 29, 2007 6:01 am

My portable CD-player has two connectors, line-out and headphone-out, the portable CD-player of my wife has only one connector, with the writing "phone/line out". Maybe they are the same at such a cheap device.

My current monitor is not good enough to see a difference, and my new one is still not ready yet. No bigger troubles, just not enough time...
User avatar
DrZarkov
I think I've had a cranial implosion.
 
Posts: 1041
Joined: Mon Feb 19, 2007 11:28 pm
Location: Kamp-Lintfort, Germany

Postby Klaas Robers » Tue May 29, 2007 6:24 am

Andrew, look at the grey value of the back ground area from left to right. I see that on both "headhone" pictures that the left part is somewhat brighter than the centre part. This is because in the centre the white circle is larger, making the mean brightness of those lines higher then at the left and right, where no white circle is present.
When really low frequencies have a lower amplitude, who cares about 30 Hz in headphones, the mean brightness of the video signal of lines is drifting towards 50% faster.
When you look only at definition (sharpness) then you might see little difference, as sharpness is in the high tones. Yes, you should learn how to look at these things.

You can do this to see the difference between line- and headphone-:
- run NBTV disc nr. 1 tracks 20-21-22, the vertical grey bars.
- connect your oscilloscope to the left line-output of the CD-player,
- switch the oscilloscope input to DC coupled.
- trigger you oscilloscope to the mains (50 Hz)
- adjust the scanning speed to see a complete frame.
- now watch variation in the bottom of the sync pulses.
- the bottom of all the sync pulses should be at the same height.
- do this too for the headphone output.
Yes, this is a tough test for a CD-player, but not impossible.

You can do this too:
- connect your oscilloscope to the R-output (sound) of the CD-player
- oscilloscope still to DC-coupled,
- run track 3 of the same CD (1 kHz at 0 dB)
- read the amplitude (peak-peak) of the sine wave
- now run track 5 (2 Hz at 0 dB)
- compare the amplitude of this slow sine wave.
- it should be between 0,7x and 1x the amplitude of the 1 kHz sine wave.
- do this too for the headphone output.
And then surprise me (or you). On my CD-players line output the amplitude of the 2Hz sinewave is about 95% of the 1kHz amplitude.
For an invisible "sag" in NBTV the -3dB cut-off frequency should be 2 Hz or lower. This will say that the amplitude for 2 Hz sinewaves must be still more than 70%.
User avatar
Klaas Robers
"Gomez!", "Oh Morticia."
 
Posts: 1656
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Postby AncientBrit » Wed May 30, 2007 10:56 pm

Lack of low frequency response in NBTV signals causes quite a few problems.

This is particularly noticeable if an image comprises say broad vertical black and white bars.

The lack of direct coupling means that the waveform ends up with a pronounced tilt at the these locations.

If the video is 70% then the tilt can exceed the amplitude of the sync pulses.

So any clamping or dc restoration circuit will act on the signal content rather than the sync signals.

This causes false syncs to be detected and the displayed image will "pull" at these transitions.

One solution that was proposed by an NBTVA member (Jeremy, I think) was to change the ratio of video from 70:30 to be more like 50:50 or 40:60.

This ensured that the sync always predominates irrespective of signal tilt.

The downside of this modification is that signal to noise ratios will suffer.

If the NBTV signal was crystal derived then crystal locking of the rasters at send and receive locations could be employed.

This was done I believe when the re-creation of the Transatlantic broadcasts were carried out a few years ago.


Regards,


Graham
AncientBrit
Green padded cells are quite homely.
 
Posts: 858
Joined: Mon Mar 26, 2007 10:15 pm
Location: Billericay, UK

Postby Stephen » Thu May 31, 2007 4:50 am

AncientBrit wrote:Lack of low frequency response in NBTV signals causes quite a few problems.

This is particularly noticeable if an image comprises say broad vertical black and white bars.

The lack of direct coupling means that the waveform ends up with a pronounced tilt at the these locations.
John Logie Baird's first successful demonstration of real television, that is, being able to reproduce moving images in continuous tones with reflected light, that he performed in October 1925 overcame the sluggishness of both the photocell pickup and amplifier system with a combination of techniques that comprised optically chopping the light to the photocell and using differential amplification to provide video peaking in the amplifier. The differential amplification "squared up" the sides of the pulse amplitude modulated pulses. See his British Patent 270,222, filed 21 October 1925 for details. See the details of his optical chopper system in British Patent 235,619, filed 12 March 1924. These patents are in the "Patents and Articles" section of the forum.

The interesting thing about using such a pulse amplitude modulated (PAM) signal is not only that it may prevent blurring of individual pixels as the display disc or drum reconstitutes the image, but in the case that the minimum black level of the signal has a value greater than a "blacker than black" value between pulses the signal need not have significant low frequency response and it will not suffer from AC coupling.

Many years ago I acquired a couple of old Western Union telefax machines that dated to the late 1940s. Interestingly, they had chopper wheels between the light source and the scanning drum to generate a PAM audio frequency signal that was within the bandpass of the Western Union lines.

It would be interesting to apply such a technique to NBTV. Generating such a PAM signal could be done by optical or electronic means. One optical technique might be to just place a sampling mask in front of the camera scanning disc that comprises a set of 48 narrow horizontal slots, in the case of the NBTVA 32 line format. Of course, the scanned image in this simple embodiment would have to have a minimum black brightness level greater than the "blacker than black" level between pulses.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Another approach.

Postby Stephen » Thu May 31, 2007 11:35 am

Mr. Baird also had another means for chopping each line into individual pixels instead of scanning continuous lines in his British Patent 236,978, filed 17 March 1924. This patent is also in the Patents and Articles section. According to the description, the example shown in Figures 1 and 2 seems to generate 12 lines of 33 pixels each at 8 frames per second.

Each slot in the radial slotted disc as shown in Figure 2 generates a line of a frame. With the radial slotted disc revolving at 1 rps and with an image sized and positioned between adjacent slots it generates 8 fps. The spiral apertured disc shown in Figure 1 does not have the purpose of scanning lines, but rather the purpose of chopping or subdividing each of the lines that the radial slotted disc generates into individual pixels. Since it has 33 apertures, this would mean 33 pixels per line. If this disc revolves at 96 rps, it would chop 12 lines into 33 pixels each per frame.

Each aperture would only be spaced by the thickness of each slot, so the apertured disc could have a very small diameter or in the alternative could have multiple spirals. If one were to adapt this embodiment to the NBTVA standard, leaving the slotted disc with 8 slots, the apertured disc could have four spirals of 48 apertures spaced one slot or line apart. The radial disc would revolve at 93.75 rpm and the apertured disc would revolve at 6000 rpm.

By the way, the image would have a position along the bottom of the disc rather than the side for vertical bottom to top scanning.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Postby DrZarkov » Thu May 31, 2007 3:22 pm

As you can see on old pictures of Mr. Bairds devices, he used it in the beginning, but later it seems that he has abandoned this idea. I wonder why?
User avatar
DrZarkov
I think I've had a cranial implosion.
 
Posts: 1041
Joined: Mon Feb 19, 2007 11:28 pm
Location: Kamp-Lintfort, Germany

Low frequency response.

Postby Steve Anderson » Thu May 31, 2007 3:46 pm

Lack of low frequency response in NBTV signals causes quite a few problems.

This is particularly noticeable if an image comprises say broad vertical black and white bars.

The lack of direct coupling means that the waveform ends up with a pronounced tilt at the these locations.


Gents, I wrote an item for both the NBTVA and the BATC for their publications about this very problem and one way of getting around it. The proof returned to me by the Editor of the BATC CQTV magazine is attached. It was published 'as is'.

Unfortunately the version that appeared in the NBTVA newsletter (Vol. 31 No. 1) contained so many errors it made the item worthless. In addition the two photos referred to were omitted.

I'm working on other ideas to circumvent the LF problem which is the biggest (electronic) problem for NBTV in my opinion.

Steve A.
Attachments
VCR Recorder.pdf
(1.68 MiB) Downloaded 712 times
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5361
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Postby AncientBrit » Thu May 31, 2007 5:45 pm

Steve,

An alternative method of recording NBTV to video tape (as opposed to audio tape) is to up-convert the signal to 625/50 (or 525/60) and record directly to VCR.

I have built a PIC based up converter where the NBTV signal is sampled directly by the ADC in the PIC.

This cuts down on chip count.


Regards,


Graham
AncientBrit
Green padded cells are quite homely.
 
Posts: 858
Joined: Mon Mar 26, 2007 10:15 pm
Location: Billericay, UK

Buying VCRs

Postby Steve Anderson » Thu May 31, 2007 6:21 pm

An alternative method of recording NBTV to video tape (as opposed to audio tape) is to up-convert the signal to 625/50 (or 525/60) and record directly to VCR.


That is one approach I did consider at the time of writing that item and I left it for later. However, my VCR broke down and it didn't seem worth getting it repaired. So I went shopping for a new one...do you think I could find one up for sale? No.

In Europe or the US there are probably specialist dealers that still have VHS machines up for sale, but not here. The machine was a JVC SVHS machine with the extended luminance bandwidth to 3.5MHz, plus it had a comb-filter decoder, so no cross-colour.

The VHS is dead, long live the VHS.

We don't have an equivalent of the UK 'Sky box' here, and disc based recorders have yet to make an appearance. I was tempted to bring one back from the UK when I was there late last year, but I already had so much stuff that I had to ship two large boxes back weighing over 120kg. I had to stop somewhere! (That's not including the 40kg allowance on the flight which I exceeded!).

So for the time being I have no way of recording FSTV, unless one considers PC video-cards. The thought of that makes me cringe.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 5361
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Postby AncientBrit » Thu May 31, 2007 10:38 pm

Steve,

Understood,

Graham
AncientBrit
Green padded cells are quite homely.
 
Posts: 858
Joined: Mon Mar 26, 2007 10:15 pm
Location: Billericay, UK

Postby Stephen » Fri Jun 01, 2007 2:59 am

DrZarkov wrote:As you can see on old pictures of Mr. Bairds devices, he used it in the beginning, but later it seems that he has abandoned this idea. I wonder why?
I suspect that once he had photocells with better response time, the extra complexity of the optical chopper wheel and its associated drive was not worth the small increase in picture quality.

In order to secure the benefit of a PAM signal that would not require DC coupling or restoration in the transmission chain, the image intensity would have had to be controlled so that black level would always be above the "blacker than black" level between pulses. This would have been difficult to achieve in practice so that benefit probably would have not been available either.

Having said this though, there probably are clever ways to achieve this result, either optically or electronically. For instance, the response of the photocell amplifier might be non-linear below a certain light level so that it has greater gain at low light levels. This would boost output of any detected light to better differentiate black image levels from the "blacker than black" level between pulses.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am


Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 33 guests

cron