Mr. ROS speaks out – an exclusive interview

Two weeks ago we explained (translated version here) how to disable the auto-spotting feature in the ROS digital communication software. Its creator, José Alberto Nieto Ros, contacted us shortly afterwards to dispute the usefulness of this operation, arguing that the DX-cluster saturation issue has been resolved. We then thought about interviewing José to know him better and give him the chance to reply to criticism on his software.

This is an exclusive interview made by Please link this article if you want share this news!

What’s your current occupation?

I’m a telecommunication engineer and I’m currently working as a consultant in the field of digital transmissions.

When did you start working on ROS? And why?

Actually, I started working in something similar to ROS when I was studying in the University, 12 years ago. I produced an  HF communications software that used the Serial Port, in a way similar to HamComm but far more robust. It ran from 50 to 1500 bauds with AM Rigs.

Last year (2009) I decided continue with the experiments, but in this case, I decided to upload the project to the Internet for Ham Radio used  because ROS offers improvements over others weak signal modes. As you can see by simply viewing the PSK-Reporter website, usage has been widely taken up by the Ham community.

Of particular note, ROS is not a time-locked mode, thus it doesn’t need a precise clock synchronization between computers, as opposed to other weak-signal modes such as WSPR. This, combined with superior data-rates and interactivity (TX and RX timing are managed by the user and it’s practical to have live-keyboard conversation just like BPSK31), would allow a DX’pedition to work in remote locations without needing accurate clock synchronization via Internet or GPS.

Are you interested at all in other means of radio-communication? Do you plan to get an amateur radio licence?

I expect get a license soon. I am interested in QRP. Also and military data com, Slow scan digital TV, DSP, in general.

How do you develop ROS? How do you test it?

Having  previously produced a working  data system, I set about improving the system making use of modern techniques, this gave me a clear objective for the ROS project, to create a mode efficient in power.

Initial testing was done with CO2DC over 7 MHz band.

Is it true that ROS is all developed in Visual Basic? Will you consider porting ROS to other operating systems?

Yes, it’s true, and not at the moment. Perhaps in a future.

How about ROS MF?

ROS MF is a mode that I created for request of Graham, G0NBD and others enthusiasts of LF and MF bands and has established itself as the mode of choice for data MF live QSOs and reactive beacons linking back to the reporting network.

It is a narrow band mode that occupies only 100 Hz according band-plans of many countries. It uses two symbol-rates, 0.8 bauds (ROS MF1) for use in Beacon Mode mainly, but  is a  live keyboard mode as is  7 bauds (ROS MF7) for keyboard to keyboard communications to 130 characters/minute.

Which are the most interesting DX QSOs made with ROS that you are aware of?

Probably across USA from Canada to Cuba with 25 mW at 150 characters/minute. A lot of Great QSOs are published in ROS website, as low power MF7 QSOs into Europe using the 500 kHz band.

What’s your recommended setup (RX/TX, antenna, computer) to make ROS contacts?

Any modern SSB radio, IF filtering is not a  requirement  as ROS  has adaptive filtering that is matched to the mode in use, AGC may be disabled, taking care not to overload the soundcard, of course.

Narrow IF filtering will degrade the system performance. ROS has been developed to run within the limitations of a simple HF transceiver.

Any antenna is suited, but with the enhanced  S/N capability of over 10 dB over the most popular data modes and its inherent insensitivity to phase distortion over the polar paths, any simple array is capable of excellent  results. Power levels of 5 to 20 Watts  are being reported as quite sufficient, the software also provides instant  S/N and fade margin reading, enabling the user to evaluate the path and power levels during the QSO, or simply by monitoring the traffic.

Computer clock must be over 1Ghz (PIII 1.2 GHz not fast enough), older PCs  can have problems using ROS. The ROS  system is leading-edge technology and it follows that demands are placed on the computer and soundcard. You don’t need any special soundcard, though.

Will you release full ROS specifications so that it could be implemented in other multi-mode programmes such as fldigi and MixW?

…………….$64,000 question. Not at the  moment, as ROS is under constant development, so it’s pointless to try and implement it elsewhere. ROS is freely available and fully web-integrated, so its not a detrimental  situation.

There has been lots of criticism over the auto-spotting feature in the ROS programme, what did you do to improve it?

Many people publish articles on the internet about ROS, without knowing how ROS actually works, which in turn have produced erroneous technical discussions and some quite bizarre actions by people otherwise considered quite sane.

There was a bug in the auto-spotting on the 15 m band, but it has been corrected quickly with the release of a new version of the software. The auto-spotting is regulated by two algorithms that compare decoded calls with the DX-cluster spots and their purpose is to limit the number of spots sent to the DX-cluster itself:

– 20 metres algorithm. It allows one ROS spot every 10 “non-ROS” spots on this band. The spot is first subjected to a “draw” among all ROS stations active on the band and with auto-spotting enabled according to three priorities:

HIGH – QRB > 4000 km

MEDIUM – call is spotted on 14.078, 14.079 or 12.112 MHz

LOW – call is spotted on 14.103 MHz

– 15 and 40 metres algorithm. It allows one ROS spot every 10 “non-ROS” spots on the sum of all bands. The “draw” is present but there are no priorities.

In the rest of bands there is not any algorithm. Spots are sent directly because activity is very low. This promotes the possibility of doing a QSO in bands like 80 metres or 10 metres where activity is very low. Algorithms can change with new versions, according the increase of ROS users in unusual bands.

It is for there reasons that I discourage to disable the auto-spotting. Now, look the DX-cluster, count the number of spots and tell me if actually ROS is saturating the DX-cluster.

11) Could you tell us what has happened in the USA with the ARRL?

USA Ham community has a “local” problem, in that, the FCC has a clause in the licence preventing spread spectrum communication below 220 MHz.

ROS is not a “SS” mode. In fact, ROS7/100 occupies 100 Hz, ten times less the Olivia32/1000 and transmits at the same rate. So, if ROS7/100 is defined as SS, then Olivia and  most of the  other  soundcard-based data modes should be classified as  SS too, or is this just  “sour grapes'” as ROS was invented outside the USA?

Totally lost in a modern sea of technicalities, the representatives of USA hams, instead of trying to amend the situation, are blindly following a minority opinion and are trying convince “all” USA Hams that all the ROS modes HF/EME/MF ROS/100 Hz are illegal and cause interferences, but Olivia 2000 Hz and other modes are legal, that’s something that I think, nobody can believe.

To state that  ROS is a “narrow-band SS”. It is the first time that I heard that a SS mode can occupy less bandwidth than others modes that the FCC says are not SS, like Olivia, MT63, MFSK, Chirp and the new super wideband ALE networks that now use the HAM bands for private e-mail exchange. It’s perhaps very clear that the ARRL doesn’t defend the interest of USA Hams. Checking PSK- Reporter for ROS activity shows my mode to be in constant use, from VHF EME contacts, through HF, to setting distance records on the LF bands, while only one user shows in the USA.

USA Hams that do not agree with ARRL should not pay the ARRL fees, as even China is free to use ROS!

As of today, the  situation remains, with only one  license to allow the use of the ROS mode issued by the FCC to WE9XLQ, a sort of cross between Eliot Ness and the Lone Ranger.

What Next?

Related Articles

11 Responses to "Mr. ROS speaks out – an exclusive interview"

  1. IZ8IYX IZ8IYX ha detto:

    Hi Andrea IZ1IVA, congratulations for the interview, thanks José Nieto Alberto Ros for choosing dxcoffee.

  2. IK8YFU ha detto:

    Congratulations for the article that explains many things!

  3. IZ1IVA ha detto:


    José Alberto Nieto Ros has sent in further comments, reflecting the latest developments in the ROS software.

    The auto-spotting algorithms are now reduced from three to two:

    – 20 metres: same algorithm as before, but it now allows 1 spot every 8 on this band

    – all other bands: 1 spot every 8 on the sum of all bands except 20 meters, The “draw” is present but there are no priorities.

  4. IZ4AKS ha detto:

    Interesting point of view expecially about the ARRL.
    At least I discover what & who is Ros.

  5. IZ8ESX Davide ha detto:

    Excellent interview, with a lot of interesting hints about what ROS is and how it works, I agree with Mr.Ros point of view about spread spectrum..
    I found the auto-spot item totally against the ham spirit, so would be better to do not selfspot at all.
    I’m not into digi-modes, but the S/N ratio of Ros-qso is just amazing.

    HAPPY DXing, Davide IZ8ESX

  6. OD5NJ ha detto:

    I like the web and I like coffee too…73′ to DxCoffee fans

  7. Frank, IZ7AUH/AK1CQ ha detto:

    great job!

  8. hb9twu ha detto:

    thank for all info is no good for qso and usa 73 greg hb9twu

  9. Colin ha detto:

    Thanks for the article, I think the s/w is good but the cluster issue is a serious one, if using this mode then the cluster option is a must! Spot only if you have found something of interest to others that maywell be a rare or the showing of bands being open. Thanks

  10. EI6IZ Brendan ha detto:

    ANY form of auto-spotting is not welcome on the cluster network, regardless of the ‘algorithm’ this is not what it was designed for. Spotting should be under user control only.

    it would be very helpful if the author of the ROS software would engage in Dialog with both his users and the Cluster sysops about how best to use the custer system in a constructive way.
    ROS does not represent 1 in every 10 QSO’s why should it be ‘ok’ for 1 in every 10 spots to be an automated ROSmode spot with an automated comment

    Brendan EI6IZ Sysop of EI7MRE

  11. Dirk G1TLH ha detto:

    As one of the authors of DX Cluster software, in fact of the preferred package that ROS seems to use, I wonder whether I could use DXCoffee to open a channel of dialog with the author of ROS.

    There is a mechanism in DXSpider that is designed for usage as a general communications channel. It is the “chat text” command. This, together with the “join ” command will allow the software a means for all ROS users to communicate with one another without alienating other DX Cluster users. Please type “help chat” when logged in for more information.

    As it stands at the moment, there is effectively a war of attrition going on between the author of ROS and a large section of the DX Cluster world. Several steps have already been taken to try to prevent the ROS mode spots overwhelming more traditional spots. I suspect the ARRL are trying to achieve the same thing via different route. If things carry on as they are further steps will be taken to reduce the effectiveness of using the DX Spot mechanism by ROS.

    It does not need to be like this. There is a perfect mechanism available – please either just use it or get in touch with me and I will help.

Leave a Reply

Submit Comment


Utilizzando il sito, accetti l'utilizzo dei cookie da parte nostra. maggiori informazioni

Questo sito utilizza i cookie per fonire la migliore esperienza di navigazione possibile. Continuando a utilizzare questo sito senza modificare le impostazioni dei cookie o clicchi su "Accetta" permetti al loro utilizzo.