Specification
Unfortunately, there are various programs out there to send and receive digital sstv. Each one has it's own specs and formats so to delve into the specs here would be useless because it would change depending on the software you use.
I am instead going to explain some of the features and program uses of the most common software used here in the states. This would be EasyPal. The software does not require outside external programs to make it work as some do. It installs from one single installation file and contains everything you need to make it work including the codec needed. A codec can be thought of as an encoder/decoder for a particular format of digital data. All digital modulation schemes use codecs, even those mp3s you are already used to using uses an mp3 codec.
Using DSSTV, the picture is broken into small blocks of transmitted data. On EasyPal, when the picture begins, you will see the sending callsign pop into the window in the upper left corner. This identifies the sending station, and is also able to be used for legal ID purposes according to current FCC rules. There are various levels of synchronization during the "hookup" process between the sender and receiving stations, they are shown beside the callsign block. The callsign is displayed during the FAC sync period. A complete picture is not decoded unless you maintain a sync at the MSC level, however.
During the header transmission, the various levels of sync will show up in your sync block moving towards the top, which is MSC. As you watch the block information group, you'll see the total blocks in the picture and the number of blocks received along with the number of blocks checked as OK and the number of the block currently being received.
The key to obtaining a complete picture is that you must have good SNR (Signal to Noise Ratio). This must be in the green level to receive a picture. I have found a signal level of S9 or higher will produce acceptable picture decoding.
The good thing about DSSTV especially with EasyPal is that if you miss too many blocks and do not get a picture decoded, you have two options to have the picture sent to you again. One is manual and one is automatic. The manual one is to ask a station which has a better signal to your QTH to just resend the picture. However, there is a BSR function on EasyPal. With this function, you can send a BSR (Block Send Request) and any station which has received those blocks can send them to you. This way you don't have to waste bandwidth and time by getting a whole picture send to you again, only the missing blocks will be resent. This makes for the best use of bandwidth as well as makes the picture take less time to appear on your screen.
One other thing to note here is that not only can you send very high resolution digital pictures on EasyPal, you can also send "waterfall pictures" and text. EasyPal will send any file you can insert into the program. Text files as well as pictures can be transmitted. And waterfall text is also available along with waterfall pictures.
Here is the receive data items you can view and see the areas I was talking about above.
Listen to EasyPal
The sound of digital EasyPal is unique to Amateur Radio, but does sound like other digital voice modes that can be heard on SW by tuning around the bands. It is sort of like a buzzing sound, but can include a swishing noise mixed with it. I'll include links to two different samples of EasyPal sounds. One is the normal picture transmission and the other is a waterfall transmission.
This is the sound of a waterfall picture.
Waterfall Picture
This is a sample from a full color picture of 20K which will take about 92 seconds to transmit. This is only a small sample from the transmission, but will give you an idea of what it sounds like.
Regular Picture
This is 61 blocks or segments of a picture containing 183 segments.
This is a tune function. The short burst you hear at the beginning is a waterfall ID send. It simply sends your callsign in the waterfall like text. This is also all that is required to satisfy FCC rules.
ID and Tune
Waterfall is used to view the incoming signal as well as display text. Here is the waterfall during a received picture over the air. You will see 3 distinct white lines in the waterfall. They will fall under the three matching tuning marks on the display. The better the signal quality, the sharper and more defined those lines will appear. Since this is an actual picture being received and the signal strength was weak at around S7, the picture would not decode, but the signal is still present. There were many missing segments on this picture for which a BSR needed to be sent to get the missing segments.
Frequencies
Probably the most used frequency where there is also something going on would be on 40 meters at 7.130 MHz. on LSB. You can also find a little activity once in a while on 20 at 14.233 to 236 MHz. on USB. With good signals this can be quite an interesting mode to operate. Pictures of all kinds are transmitted including family members, pets, even cartoons and I have seen people sending newspaper articles and cartoons for all to see. I personally have created 3D pictures and sent them as many hams have 3D glasses in their shacks. Whatever you can create can be sent with lots of talk and rag chewing between pictures. The gang on 7130 are very friendly and will help you with your software no matter what you are running.