What is the difference between a segd recorded data and a segy recorded data and can we convert from segd to segy and vice versa?
SEG-D is a specialized format, while SEG-Y is a general-purpose format.
In general, SEG-D is intended for field recordings of seismic data, and SEG-Y is intended for 'seismic data exchange'. Having said this, SEG-Y is so general-purpose, and so ubiquitous, that I'm not surprised to hear (anecdotally) that people are using it for data acquisition. Certainly, the recent release of Revision 2 — which allows for high sample rates, variable trace lengths, long records, and large files — facilitates this.
As an interpreter, I use SEG-Y all the time, whereas I've never come into contact with SEG-D. Thus, I don't know enough about it to go into any detail, but a quick look over the SEG-D specification suggests it has better support for describing receivers, channels, sources, and including things like observer logs. Which is what you'd expect in an acquisition format.
SEG's formats are notoriously, er, flexible. So converting directly between the two formats may or may not be possible, depending on which fields have been used, and the lengths to which you're prepared to go to 'adapt' the target format to your use case. The usual way to convert, as it were, would be to process (i.e. demux, sort, image, and maybe stack) the field data in the SEG-D file(s), and write out one or more SEG-Y files.