[Parabola]

Thanks Tas and Jeffrey, more fivers to post. That's the entire marketing budget spent now. I knew I should've allowed at least £20, but I spent it all on PROMs

Tas, I still haven't fixed the high frequency grating in the 625 TCC source image, but the version I have now is nicer than what you currently have in that slot. I can always post you a PROM so as to avoid the Christmas traffic.

Originally Posted by Amraduk

.....

Although both are no longer available, Richard Russell and Dave Grant's testcard generators also supported teletext!

As far as I'm concerned, although I've never owned one or even held one, Richard Russell's generator is the gold standard in this class of device and nearly everything I have done is to try to go some way to emulating what is possible with his design. I'm getting there, but IMOGen still needs a "next generation" to become easily user programmable, and I've not even attempted to add some of the extra features that his unit can offer but are not of so much interest to the vintage tv user (widescreen signalling and fast text, for instance).

I think that Jeffrey was being kind to me here by promoting the fact that, in it's current form, IMOGen can store up to six text pages. This is more than the four that RTR's generator stores at one time. This is related to the available PROM space and the number of complete frames stored. The other feature that I incorporated is the front panel selectability of the VBI signals. This allows the instant blanking of the Teletext data lines and VITS if required. Of course, you can choose whether to program them, or not, with RTR's generator, but that does involve a trip to a PC to achieve the same result.

As far as I know, and I know very little about what was offered with Dave Grant's Dinosaur TCGen, it only produces one page of text which contains some design details and credits about the generator itself. I'd be interested to hear differently because it should certainly be able to do at least four if correctly coded to. I don't believe the VBI signals would be blankable without reprogramming the PROMs.

Also what Jeffrey may be thinking of is my intention to try to double the available text pages with a new version of firmware in future. This will involve some address swapping manouvures during the relevant data lines and should be possible as long as I don't run out of macrocells in the CPLD to implement the logic. There is plenty of space in the PROMs. I need to get my VHDL head on (watch out!) to look at this seriously in the near future.

As you say, the main problem with the two excellent designs that you mentioned is that, nowadays, you cannot buy them anymore. Hence my efforts with IMOGen, mainly to satisfy my own need.

Jeffrey has also been kind by not mentioning that our recent comparative tests showed that the reconstruction filter in IMOGen is not quite as good as the Dinosaur design. This isn't all that surprising, because the Dinosaur filter is more complex and looks like it's been designed by someone who knew what they were doing, whereas mine is relatively simple and was designed by some freeware analogue filter software. This is the most tricky and infuriating element of the design, as trying to make a half decent reconstruction filter, I've learnt, at a sample rate as low as 12MHz is always going to be a series of compromises. I'm still working to improve this at the moment, even though the differences we are talking about can only really be discerened on a professional monitor and are, mostly, not an issue with real world vintage TVs.