It is currently Sun Nov 19, 2017 9:49 am

PIC 405 line pattern generator

Valves, Transistors, IC's and Components. Bench test techniques, Test Gear - Valve Testers, Scopes, CRT Testers, Bridges, Sig Gens, Pattern Gens, Workshop tours etc
 
Posts: 1931
Joined: Mon Jul 25, 2011 12:12 pm
Location: Burghfield, Reading, UK

Re: PIC 405 line pattern generator

Post by Brianc » Thu Feb 11, 2016 8:15 pm

The lineup sawtooth was linear as it was to test the linearity of the distribution system, not the acquisition or display systems. In every source cubicle (say TK or VT) there was a locking Kellog key which sent a 1V sawtooth and zero level 1kHz tone to line. The levels were always checked in CAR and then at the destination (studio or pres). Non of this chroma phase milarky!

 
Posts: 692
Joined: Sun Aug 07, 2011 8:48 am
Location: Edinburgh

Re: PIC 405 line pattern generator

Post by peter scott » Thu Feb 11, 2016 11:59 pm

Brianc wrote:You're right (again) Cathy. I was getting mixed up with the terms. In fact the term we used was just "bars" (I never heard them referred to as "Academy bars"), These were used a lot as a general purpose test signal (hence my confusion with pulse cross). The line-up signal, of course, was sawtooth. - "Sawtooth and tone on OS1".


"Art Bars?"
http://www.transdiffusion.org/2003/08/23/tvreturn

Peter

Image

 
Posts: 2714
Joined: Sun Sep 01, 2013 6:19 pm
Location: Behind the sofa

Re: PIC 405 line pattern generator

Post by Cathovisor » Fri Feb 12, 2016 12:07 am


 
Posts: 1931
Joined: Mon Jul 25, 2011 12:12 pm
Location: Burghfield, Reading, UK

Re: PIC 405 line pattern generator

Post by Brianc » Fri Feb 12, 2016 1:27 pm

In case you're wondering where the word "Academy" in my earlier post came from, it must have been lingering in my mind, left over from the aspect ratio discussion. I seem to be very good at reading what I want to see rather than what's actually there! :zx:

 
Posts: 1931
Joined: Mon Jul 25, 2011 12:12 pm
Location: Burghfield, Reading, UK

Re: PIC 405 line pattern generator

Post by Brianc » Fri Feb 12, 2016 1:56 pm

I've just read the original article in the 1946 BBC Year Book. Funnily enough, it doesn't mention "Art Bars" and the picture of the TV studio is different from that shown in the later article. However, on thinking about it, the word art could well be an abbreviation of the word artificial and I have definitely hear that term.
Incidentally, the dust jacket for the 1946 year book is very evocative of WWII ending - here it is:
1946-yearbook.jpg

 
Posts: 2714
Joined: Sun Sep 01, 2013 6:19 pm
Location: Behind the sofa

Re: PIC 405 line pattern generator

Post by Cathovisor » Fri Feb 12, 2016 3:06 pm

It is indeed a contraction of 'artificial' - I have the feeling it has a parallel in German too; for some reason Kunst- comes to mind? I think you can have something like Kunstakademie and Kunstleder...

I have that Year Book too - I also have a 1945 Bush catalogue entitled "Over To You" which similarly uses a split image; the black/white half of the scene representing war whilst the colour half represents peace.

 
Posts: 1931
Joined: Mon Jul 25, 2011 12:12 pm
Location: Burghfield, Reading, UK

Re: PIC 405 line pattern generator

Post by Brianc » Fri Feb 12, 2016 4:11 pm

One of my little treasures is a complete set of year books from 1928 to 1978. After that date, I lost interest. Two were, of course, not published - 1953/4. I'm not sure why not but it may have had something to do with QEII's coronation and the expense of the BBC's coverage! Does anyone know for sure why?

 
Posts: 185
Joined: Sat Jul 23, 2011 10:40 pm
Location: Roscommon Ireland

Re: PIC 405 line pattern generator

Post by FRANK.C » Sun Feb 14, 2016 7:29 pm

I have added another test pattern.
This one will give an indication of how good or not the grey scale tracking is between dark and bright scenes.

A nine step grey scale is in the center of the picture.
The background starts off at black level, every second the background increases in brightness by one grey scale step until it reaches peak white, it then reverts to black level. That sequence keeps repeating itself.

Frank
Attachments
testcard28.jpg

 
Posts: 1931
Joined: Mon Jul 25, 2011 12:12 pm
Location: Burghfield, Reading, UK

Re: PIC 405 line pattern generator

Post by Brianc » Sun Feb 14, 2016 7:43 pm

This raises an interesting question. In my broadcasting days, the transmitted signal had to be gamma corrected by stretching the blacks to compensate for the triode-like transfer characteristic of the CRT. This gives an overall linear light transfer from scene to CRT picture. Is this still done as the transfer characteristic of the modern digital display can be made linear or whatever law is desired and there are so very few CRT TVs in use today?

 
Posts: 2714
Joined: Sun Sep 01, 2013 6:19 pm
Location: Behind the sofa

Re: PIC 405 line pattern generator

Post by Cathovisor » Sun Feb 14, 2016 8:08 pm

Yes - TV cameras are gamma-corrected to approx. 0.45 although these days it can be user-selectable for effect. Back in the days when I was a trainee we had a unit known colloquially as the GNU - the Gamma Neutralising Unit, which was used to give a certain effect to pictures.

Although there are an increasing number of professional OLED monitors in use for 'Grade 1' applications, in some places you can still see CRTs doing that job - usually the very nice Ikegami 20" ones (I have one, I case I ever want to go HD here at home :qq1 )

It's a mistake to assume that nearly everyone has gone down the flat panel path - I know plenty of ordinary people with CRT sets still; the occasional watchers of TV like myself who see absolutely no reason to buy one of these oversized, abysmal-sounding monstrosities.

 
Posts: 3028
Joined: Thu Jun 28, 2012 12:20 pm
Location: The Vale Of York

Re: PIC 405 line pattern generator

Post by Katie Bush » Sun Feb 14, 2016 9:05 pm

Cathovisor wrote:
It's a mistake to assume that nearly everyone has gone down the flat panel path - I know plenty of ordinary people with CRT sets still; the occasional watchers of TV like myself who see absolutely no reason to buy one of these oversized, abysmal-sounding monstrosities.


Definitely second that (e)motion..... :aal :aad

Marion

 
Posts: 2714
Joined: Sun Sep 01, 2013 6:19 pm
Location: Behind the sofa

Re: PIC 405 line pattern generator

Post by Cathovisor » Sun Feb 14, 2016 9:35 pm

Katie Bush wrote:Definitely second that (e)motion..... :aal :aad

Ooh, she's gone all Smokey Robinson on me...! :qq1

 
Posts: 185
Joined: Sat Jul 23, 2011 10:40 pm
Location: Roscommon Ireland

Re: PIC 405 line pattern generator

Post by FRANK.C » Mon Feb 15, 2016 8:54 pm

Brianc wrote: In my broadcasting days, the transmitted signal had to be gamma corrected by stretching the blacks to compensate for the triode-like transfer characteristic of the CRT.

That's interesting I had assumed that sort of compensation would have been done at the receiver with the transmitted signal been linear. I guess when all receivers needs the same compensation it is easier to do it before transmission.

Frank

 
Posts: 1931
Joined: Mon Jul 25, 2011 12:12 pm
Location: Burghfield, Reading, UK

Re: PIC 405 line pattern generator

Post by Brianc » Tue Feb 16, 2016 10:54 am

Also, Frank, a gamma correction of 0.45 requires quite a lot of black stretching and amplifying noise in the blacks can be a problem so it is better to do it as soon as possible to process as quiet a signal as possible (say).

 
Posts: 2714
Joined: Sun Sep 01, 2013 6:19 pm
Location: Behind the sofa

Re: PIC 405 line pattern generator

Post by Cathovisor » Tue Feb 16, 2016 1:53 pm

... and these days that also helps any quantisation effects...

 
Posts: 185
Joined: Sat Jul 23, 2011 10:40 pm
Location: Roscommon Ireland

Re: PIC 405 line pattern generator

Post by FRANK.C » Tue Feb 16, 2016 2:44 pm

Ah its only now that the penny has finally dropped as to what this means for a pattern generator, a correct grey scale would be nonlinear, which this one I am building is linear, would it therefore be worthwhile sticking a nonlinear amp on its output or is that going OTT.

Frank

 
Posts: 63
Joined: Sun Dec 18, 2011 1:46 pm
Location: Hertford

Re: PIC 405 line pattern generator

Post by jjl » Tue Feb 16, 2016 3:31 pm

Frank

The classic method of applying a gamma transform in a digital video system would be to use the uncorrected digital pixel values as an index into a lookup table containing the gamma corrected values, the corrected digital values could then be fed to a DAC. Typically, pixel data would be represented as 8 or 10 bits per colour channel at this stage; your system being monochrome has a single channel.
With your system, I suspect that using this method might get close to the limit of CPU performance and / or memory space available.

Having said that, your system effectively represents each pixel as a 4 bit value which is fed to your resistor chain DAC at its output, so there is very little scope for applying a gamma transform; only 16 different pixel values can be represented.

I see that your PIC has enough pins to provide an 8 bit output, but if you chose to pursue this, you may run into performance issues.

John

 
Posts: 2714
Joined: Sun Sep 01, 2013 6:19 pm
Location: Behind the sofa

Re: PIC 405 line pattern generator

Post by Cathovisor » Tue Feb 16, 2016 5:25 pm

For test and diagnostic purposes, we want a linear signal as currently supplied from Frank's generator - a gamma corrected signal is completely unnecessary. Certainly all test signals I've ever encountered in broadcast are linear, except when a linear sawtooth/staircase is injected into the input of a camera head amp so you can observe the transfer characteristic when gamma correction is applied.

 
Posts: 1931
Joined: Mon Jul 25, 2011 12:12 pm
Location: Burghfield, Reading, UK

Re: PIC 405 line pattern generator

Post by Brianc » Tue Feb 16, 2016 6:11 pm

Cathovisor wrote:For test and diagnostic purposes, we want a linear signal as currently supplied from Frank's generator - a gamma corrected signal is completely unnecessary. Certainly all test signals I've ever encountered in broadcast are linear, except when a linear sawtooth/staircase is injected into the input of a camera head amp so you can observe the transfer characteristic when gamma correction is applied.
I totally agree, Cathy. The point that I was making when I brought up the question of gamma was that the linear stairstep Frank is generating would not indicate of how good or not the grey scale tracking is between dark and bright scenes because of the 2.5 gamma of the CRT. Were gamma correction be applied, then a light meter would be needed to measure the linearity of the brightness steps so a bit pointless, really!

 
Posts: 185
Joined: Sat Jul 23, 2011 10:40 pm
Location: Roscommon Ireland

Re: PIC 405 line pattern generator

Post by FRANK.C » Tue Feb 16, 2016 7:46 pm

Thanks Chaps for clearing that up.

John
You are correct in your assessment, the PIC is close to its limits performance wise and a look up table etc. would not work.

I am only using 9 of the possible 16 pixel values. Peak white is 1000.

I was thinking about this and If Gamma correction was needed then as I am only using 8 steps above black level a single bit could be used to represent each step with its resistor weighted accordingly. Some software would need to be rewrote but I think I could get away with without rewriting a major amount.

Brian
I am probably not explaining things very well and using the wrong terminology.
Referring to the last photo I posted if the grey scale on it is set up correctly when the background is at black level, then when the background changes to peak white, I would have expected the grey scale to display the same as it did when the background was at black level (or maybe I am expecting too much). This is not the case on the TV22 that I tried it on, by the time peak white is reached the darker three steps has merged into the same shade of black/grey. I guess this it to do with the higher demand on EHT as much as anything else.

Frank

 
Posts: 1931
Joined: Mon Jul 25, 2011 12:12 pm
Location: Burghfield, Reading, UK

Re: PIC 405 line pattern generator

Post by Brianc » Tue Feb 16, 2016 8:45 pm

I would think that that effect is more to do with the video amplifier than EHT. When you think of the difference in energy of the waveform between a black background and a white one, Both amplitude linearity and DC restoration are being put under extreme pressure. Have you put a 'scope on the CRT cathode to look at the waveform? Hopefully you can use DC if the voltage is not too high - I'm having problems with my GEC project due to sag on the video amplifier anode sweep waveform.

 
Posts: 2714
Joined: Sun Sep 01, 2013 6:19 pm
Location: Behind the sofa

Re: PIC 405 line pattern generator

Post by Cathovisor » Wed Feb 17, 2016 12:37 am

Here's the video output circuit of the early TV22.
TV22 video output.PNG
TV22 video output.PNG (17.59 KiB) Viewed 1885 times

The cathode of the CRT will receive a DC potential dependent upon the video waveform via R18/20/21, but unfortunately that 0.1µF cap (C21) will also have an effect upon the picture depending upon the mean level of the vision signal. This is the effect you're seeing. Were the video to be fully DC-coupled from the detector, you wouldn't see this problem - especially if there was a proper black level clamp in the video circuit too. That however, costs money! On sets where the video is almost entirely AC-coupled (like in some dual-standard sets) then this test signal would be a really nasty one to apply. It also explains why test cards have an APL (average picture level) of about 50%.

Brian may well remember two step PLUGE and why eventually it became four step PLUGE :qq1

(PLUGE is explained here - http://www.root6.com/broadcast-engineer ... libration/)

 
Posts: 3781
Joined: Sun Feb 12, 2012 7:43 pm
Location: North Hykeham, Lincolnshire and Ilford, Essex (but not for much longer ...!)

Re: PIC 405 line pattern generator

Post by Terrykc » Wed Feb 17, 2016 11:00 am

I'm sure I once read somewhere that there had never been a specification for Test Card C. We didn't have TV in our house until I started work but I do remember well Greys in the High Street with its front door recessed a l-o-o-o-ng way back from the street with umpteen sets on each of two shelves lining the right hand side of this 'avenue'.

The sets on one shelf were tuned to BBC and those on the other to ITA. Of course, in the late 50s and early 60s, all of the sets displayed Test Card C for most of day and what still sticks in my mind is one shelf of dark grey test cards  (BBC) and the other shelf of light/mid grey test cards (ITA). This suggests to me that the ITA used gamma correction on the test card whereas the BBC didn't.                        

 
Posts: 3781
Joined: Sun Feb 12, 2012 7:43 pm
Location: North Hykeham, Lincolnshire and Ilford, Essex (but not for much longer ...!)

Re: PIC 405 line pattern generator

Post by Terrykc » Wed Feb 17, 2016 11:07 am

SHADES OF GREY

As an aside to this, the town I lived in was Grays, in Essex. The aforementioned Greys shop was almost directly opposite another TV/Electrical shop on the other side of the High Street. Its name was Sidney Gray ...

 
Posts: 185
Joined: Sat Jul 23, 2011 10:40 pm
Location: Roscommon Ireland

Re: PIC 405 line pattern generator

Post by FRANK.C » Wed Feb 17, 2016 1:58 pm

Brian and Cathovisor
Thanks for that explanation, it all makes sense.
The bench is a bit crowded at the moment but when I get a chance I will put the scope on its cathode.

Terry
That's interesting, maybe the new kids on the block wanted to show they were different.

Frank

PreviousNext

Return to In The Workshop



Who is online

Users browsing this forum: No registered users and 1 guest