Hi Paul,

not sure whether you have already an answer or found a solution ...

According to the manual ch 6.2 all internal jumper are for other purposes, so I fear there is no normal way to achieve this.
The easiest solution might be, to invest into another computer case which has more room for more slots.

Reason: even if the AES signal would be directly connected to the connector for the ribbon cable between main and daughterboard and a skilled hobbyist could build something like a small adapter, I would not recommend such a thing. That would be a loose connection without any strain relief. Then only someone would have to accidentally pull too hard on the AES cable or trip over it and then you get mechanical pull on the card. This could destroy everything mechanically or by short circuits.

Such a tinkering solution would certainly cost a lot of time and money to build a workaround with some risks involved ..

Therefore, in my opinion, the easiest thing to do would be to convert the computer into a larger housing that offers you enough free slot capacity. Then you could use not only AES but also the 3rd MADI bus, if you should ever need it in the future.

Indeed Daniel, good point.
Paul, is this for an Apple system?

According to this thread you seem to use Apple, correct?
         https://forum.rme-audio.de/viewtopic.ph … 62#p159862

Maybe you could get an external thunderbolt based expansion cabinet from Sonnet:
        https://www.sonnettech.com/product/sonnet-products.html
I think they have a good name in the Apple world ?!

Guys, thanks for your messages. Yes it’s for an apple system - I’m a Logic user Mainly and a bit of Pro Tools - currently a Mac Pro 5,1 but I expect to change to the new mac pro with apple silicon when release later this year or next.

The Mac Pro I have has 3 slots but 2 are already used with UAD-2 cards and I have checked there is no room for daughter cards and really I am not so concerned about now but rather if it will do well in my next studio machine. If the MADI FX had been a good fit now I would have bought it. Glad I didn’t though.

Since I posted this I found a great price on a ADI-648 and then straight away a Madiface USB. So now the Madiface USB has replaced the original UFX as interface feeding (so far) 2 x cranborne Audio 500 Adat racks and the old UFX as headphone sends and ADI-2 Pro FS BE as control room and mixdown.

The big payoff for me with this system is still having enough IO running the whole system at 96Khz. The RTL of Madiface USB is about the same or bit better than UFX minus converter latency so I’m happy for now. Now that I have gathered this new gear maybe with next mac I may even go for UFX+ (current mac doesn’t have thunderbolt) and avoid the whole PCIe thing or stick with Madiface USB since it’s rtl is ok (about 2.9ms at 32 buffer 44.1Khz and Cranborne 500Adat ADDA)and it has more than enough IO with all my converters.

Ramses according to your own tests the RayDat had slightly better rtl than HDSPe Madi FX on your windows machine but UFX+ had the best when converter latency was added in. That’s what has me thinking about the UFX+ later on. It’s more IO than I need but it’s not that much more expensive than the HDSP MADI FX and if I had to add the cost of a Sonnet explanation chassis or similar just to run MADI FX then the price difference is almost nil.

Should also add that over on this thread < https://www.gearspace.com/board/music-c … e.html> RME’s own AIO Pro is the top performer measuring not just rtl but also CPU load. So some preform better RTL but at more CPU load so not as useful. Given the AIO Pro is the most recent release in the HDSPe range this presents that RME is still making gains in this area. It makes me wonder what the plain HDSPe Madiface would perform like in this kind of test given the RayDat did well but being an older release card.

It’s interesting RME support says technically the daughter card does not have to be inside the machine. I wonder if organising the right ribbon cable would be hard?

Thanks again for the thoughts guys...

Thanks for your kind words Ramses.

Sorry you’re right about the MADI FX being slightly faster RTL. The thing that got me was that the input side was faster on RayDat but MADI FX is quicker on the output side.

Regarding Totalmix FX the only thing I’ve ever used compression or eq on was my talkback mic. The reverb and delay is only very occasionally useful if I have to overdub in a session where a lot of plugins are already there but can live without it. I have an old lexicon lying around if I need that. And one MADI IO or as I think of it 32 IO channels at 96Khz is already more than I need - so weighing up whether the UFX+ or MADI FX is a better fit I think the one area the MADI FX might win is slightly lower CPU overhead at same RTL. Maybe because for that active load control you point out. I don’t even know for sure but it still seems the PCIe cards score better in this area than even the Thunderbolt card - even though technically PCIe and TB are meant to be the same thing.

Other than those thoughts everything else between UFX+ and MADI FX seem quite equal. Both offer the AES IO, but the one thing the UFX+ has is extra outputs including 2 x phones and other TRS outs to drive further headphone sends. I commonly need 3 - 5 headphone mixes at the same time I need all my nice inputs for recording on the 500 racks so while those racks have nice headphone amps they require other analog outputs to feed them.

Anyway it’s all a bit academic now until Apple release these new computers. Once that happens I can decide but at least now I can just keep working and now at double speed sample rates.

Thanks for all your useful comments Ramses on this and other threads. It’s been quite complex for me to work through this but it seems like I’m set for now:-)