In the late 1970's I fabricated an Analogue "monochrome" SSTV monitor.
I had purchased a pallet of ex-ABC TV studio video equipment at an auction.
I reworked an EMI waveform monitor as my construct.
I had to procure a P7 (long persistence) green phosphored 5" CRT (5BP7)
(radar display tube) in the U.K. for the display.
I later disposed of all the analogue SSTV set-up to another "experimenter" and haven't seen it since.

I used an Ian Pogson ELECTRONICS AUSTRALIA magazine Design from May 1974.

Article page 1
Article page 2
Article page 3
Article page 4
Article page 5

Another idea I tried was a slope detector using OP-AMP I.C. filters (the Robot 70 circuit)

Here is a another very similar circuit from ARA magazine 1981
(Amateur Radio Action Aust)

An alternative to "slope - detection" is the pulse counting type detector.
Although untried by myself; the RSGB (UK) handbook gave a good assurance of the circuit shown.
I have recently used the pulse-counting type detector in an FM tuner design, where FM input is 20-250KHz.
This circuit is of course configured for the 1-3 KHz region
A pair of one-shot pulse generators or monostables(74121) are triggered by the rising AND falling edges of the incoming audio sine wave.
Then the pulses are diode "or-ed" and integrated and filtered to give an amplitude o/p voltage proportional to the freq' mod input.
In the U.S. this seems to be known as a "charge-pump circuit"
Monostable circuits need a certain time to recover after triggering.
If the recovery time is not completed, the next cycle might be shortened, and an inaccuracy or jitter can result.
As a general rule , monostable circuits operate best if the recovery time exceeds the ON time, although durations of 90% of total time can be acheived.
{Source: TTL cookbook, Lancaster pg 187}

Freq Period half period pulse width duty cycle max permitted duty cycle
2300Hz 434uS 217uS 134uS 31% 67 - 90%
1500Hz 666uS 333uS 134uS 20% 67 - 90%

I must have neglected to make a note of the LC circuit values in the low-pass o/p filter.

I also tried a "digital" type sync-pulse detector

Then I designed a lockable sync pulse generator (using a CMOS 4046 PLL)
This provided a regular sync pulse string to the scanning raster should the input signal fade (as in off-air reception).
When sync signals were detected the PLL would lock onto the pulse train.

SSTV Specification 60Hz areas 50Hz areas
Horizontal line rate 15Hz (60/4) 16 2/3Hz (50/3)
Horizontal line time 66.67mS 60mS
Horizontal lines 120 120
Vertical Frame rate 1/8Hz 1/7.2Hz
Horizontal sync pulse width 5 mS 5 mS
Vertical sync pulse width 30 mS 30 mS
Sync subcarrier Freq' 1200 Hz 1200 Hz
Black Frequency 1500 Hz 1500 Hz
White Frequency 2300 Hz 2300 Hz

The results were really quite poor ; looking at images "stored" on a long-persistence phosphor CRT in a darkened room is a far cry from the full colour PC processed systems we enjoy today!

These are colour SSTV pictures ("Robot 36" 320 x240 format) received from the ARISSAT-1 satellite orbiting in space (Sept 2011)