Page tree
Skip to end of metadata
Go to start of metadata



Can the MIC Pre-Amps gains be controlled by using remote protocols HUI/Eucon/OASIS?

Right now, the MIC Pre-Amps' remote is accessible from the HORUS/HAPI Web Access, from the Pramix Mixer and through Midi Pre which makes is supported for remote controls with application such as ProTools. That is true for any connection (even Pyramix).  Remote control over Oasis/Eucon/HUI etc will be integrated later.

Is the MIC Pre-Amp in the HORUS/HAPI able to handle the DPA 130V MIC systems?

No, HORUS/HAPI supports 48V MICs.

Are the AD modules available in Line Level only (without the MIC Pre-amplifier)?

No, they are available in Mic or Line (switchable). All Analog Input modules are all switchable (on a per-channel base) between Mic and Line. It is not an option but a feature. 

Does HORUS/HAPI have an AD circuits that differ from the Sphynx II ones? Will I get different sound if I will put the Sphynx II to HORUS/HAPI the digital input and compare? 

Yes, HORUS/HAPI has an entirely re-designed Analog as well as AD conversion circuitry. In terms of "sound quality", Mic/Line and AD will be quite close between the two generations of products (all specs within 1-2 dB of THD+N and noise specs, with HORUS/HAPI being typically better. 

What happens if a MIC signal is split passively to HORUS/HAPI and a Mixer and both generate 48V? Will it blow up?

No, The HORUS/HAPI MIC inputs are designed so that if an external source also provides 48V to the input with the 48V supply of the HORUS/HAPI activated, it will not blowup.

Could we have 48 MIC Pre-Amps channels in one HORUS (6 AD modules)?

Yes, for more than 3AD modules refer to the Redundant/Secondary Power Supply requirements.

Does the HORUS/HAPI AD8 Mic Preamp provide a direct PreAmp output without latency, with parallel recording output for the monitoring operation ? Is there any possibility for this option ? 

Merging's current HORUS/HAPI AD8D and AD8DP modules provide direct preamp Analog outputs. Previous AD8 models do not have this feature, contact your Merging Sales partner fro further details.
You can also route the AD converted channels internally to the HORUS/HAPI to any number of parallel digital outputs simultaneously, including AES, MADI (Coaxial AND Optical) and of course RAVENNA with close to zero latency.

Is there a difference in audible quality between the Mic and Line input settings of the HORUS/HAPI AD8 / AD8P cards? 

We would not expect that a user would normally experience any audible shift in quality between Line and Mic modes. There are some minor changes in the Frequency response, depending on the gain settings of each, which can be measured. But, those will occur in frequency bands either above 100kHz or below 10Hz, thus way outside the audible range. 
As an addendum to this point, since the Line input must be attenuated to cope with signals coming in at levels above +13dBu, which is what is presented to the AD converter for 0dBFS, this adaptation process uses resistive division, which will in turn slightly higher the noise floor in Line mode due to the thermal resistor noise. Thus, if you can be assured that the Line level signal presented to the analogue inputs will not exceed +13dBu, there is then a very slight advantage in setting the input to Mic mode instead. Similarly, we would not recommend to engage the PAD in Mic mode, if not required, for the very same reason.

I was surprised to see that analogue out from the HORUS/HAPI comes back on analogue input less 6dB. Why? 

Merging made sure that nobody can overdrive its Line inputs (capable of accepting up to + 24 dBu for 0 dBFS) nor overdrive their existing Analogue equipment when connected to HORUS/HAPI Line Outs (+ 18 dBu for 0 dBFS) with Default ex-factory settings.

The DA8 and DA8P modules feature both hardware level settings and a software fine adjustment to align the Analog Output levels to whatever local/organization operational levels are mandated.
On DA8/DA8P cards from run 7 upwards, the hardware level setting is done via software through the option “max output level” in each DA’s setting page, allowing either + 24 dBu or 18 dBu max level.
On DA8/DA8P cards prior to run 7 the hardware level setting is in the form of 4 DIP switches per output channel (refer to the Horus or Hapi User Manual for more details)
The Hardware settings will usually be set only once, at product installation, and only if the desired Operating Line Level differs from the default ex-factory settings of +18 dBu for 0 dBFS.


Does HORUS/HAPI have a DA circuit that differs from the Sphynx-2? Will I get different sound if I will put Sphynx II to HORUS/HAPI digital input and try to compare?

Yes, HORUS/HAPI has entirely re-designed Analog as well as DA conversion circuitry. In terms of "sound quality", Line Out specs will be vastly superior with the HORUS/HAPI. 

What is the possibility to playback through the HORUS/HAPI DA converters in single ended as well as in balanced modes?

The HORUS/HAPI Line Out circuitry are fully symmetrical, but non-floating. So driving asymmetrical (single-ended) loads is easy and possible and should be wired in such a way to just use one of the symmetric outputs (either Hot or Cold) and Ground. It is not recommended to short the other (unused) output to Ground . This will also reduce the output signal level by 6 dBus compared to the full differential swing of the symmetric output, which is usually too high for usual single-ended loads. Further reduction of the level might be further required to avoid clipping the single-ended receiving side.
(for details, refer to the HORUS/HAPI User Manuals)

What happens with the HORUS/HAPI, if a sound engineer accidentally turns on his Phantom Power on the mixer which is connected to the HORUS/HAPI Output?

The HORUS/HAPI are protected against such a case. It is not recommended but the HORUS/HAPI DA output supports 48V, if driven through 6.8 kOhms resistors, as typically used in Phantom power designs.

Are the DA outputs DC coupled?

All of the Merging DA8 outputs are  DC Coupled.

HORUS/HAPI Network Accessories and Features

On one RAVENNA network, is it possible to have one HORUS/HAPI running at 44.1kHz for example and the other one at 96kHz and use them both for different DAWs or in turn?

RAVENNA allows for concurrent multiple Sampling rates on the same network, but :
While RAVENNA does indeed provide the possibility to simultaneously convey signals at different sampling rates over the same network, HORUS/HAPI are however designed to operate at a given sample rate and will only accept RAVENNA streams that are presented to it at the proper matching sampling rate. This said if, during this summer, we find out there is some FPGA real-estate left, Merging may add a few channels of real-time SRC in the architecture of the internal HORUS/HAPI router to provide such capability on a limited number of audio channels. 

Is the Merging MassCore NIC card (NET-MSC-GBEX1) temporary and why do you need this special card?

The MassCore Ethernet PCIe Network card is required for performance issues. MassCore has a dedicated driver for this (and only this) NIC. It took several months of engineering to write such a low-level driver for a specific NIC.
This said, RAVENNA (and HORUS/HAPI) will also work well without the need to use this specific card, when operating with ASIO/CoreAudio (VAD) driver.

Do I have to pull out the Mykerinos Hardware (PCI MB2) before I install the "Masscore Single port Gigabit Desktop Adapter"?

No, but be aware that in RAVENNA/MassCore only the MassCore PCIe Ethernet card will be enabled, users cannot use their Mykerinos card simultaneously. You must keep the Mykerinos card in your system if your keys are on it and if you want to switch back to Mykerinos/MassCore.

What parameters does the network need to be compatible with RAVENNA? Does it need to support some Layer 2 or 3 or Jumbo packets for example?

The Network must be Layer 3 compliant and must be a Gigabit network. For additional information please consult the following page and related ones under:

What network equipment does Merging specify for use with the RAVENNA network? Do I need to buy insanely expensive switches etc?

The types of switches and equipment you purchase relates to the size and complexity of the RAVENNA network.
A Managed network switch is required.
The Switch needs to be configured prior to usage.
See all the details in the RAVENNA / AES 67 switches page 
. For a list of network requirements, please see the RAVENNA webpage.

Does the RAVENNA PCIe card NET-MSC-GBEX1 have a secondary port or will I need to buy two such cards for redundancy?

RAVENNA redundancy is not yet implemented on the MassCore, or Core Audio side, only one ethernet adapter per computer can be used for the moment.
ASIO ST2022-7 redundancy is available since the Merging RAVENNA ASIO driver V13.04 / Merging Audio Device driver v1.0

Does the RAVENNA secondary (redundancy) port on the back of HORUS/HAPI works ?

RAVENNA redundancy is not implemented on the Horus / Hapi, only the first Ethernet port can be used.

Does the RAVENNA secondary (redundancy) port on the back of HAPI MK II works ?

The secondary port on Hapi MK II (and Anubis SPS) can be used in redundancy mode (Multi Interface - ST2022-7) or switch mode (to connect a single Merging RAVENNA / AES67 device).

How would the latency be affected if HORUS/HAPI (in RAVENNA mode) is connected via a 54mbps wireless network, assuming you have the whole bandwidth available for the HORUS/HAPI application alone, with a physical distance of 50m?

RAVENNA is not designed to transmit audio on Networks without minimum QoS requirements. Typically for now we do not recommend using RAVENNA audio streams over Wireless. However control data has no problem operating over Wireless. 

Do I need to use the Merging Ethernet card (NET-MSC-GBEX1)? I want to use my Laptop to run Pyramix with HORUS/HAPI!

It is important to realize that the MassCore Ethernet PCIe Network card will only be required when using MassCore. An ASIO (PC) and CoreAudio (MAC) driver is provided, allowing the use of HORUS/HAPI with standard PC/MAC Ethernet ports (on laptops as well!). 
The NET-MSC-GBEX1 board manufactured by Intel has its firmware reprogrammed by Merging for RAVENNA/MassCore acceleration, so this card must absolutely be supplied by Merging!

Is there any distance limitation for RAVENNA mode use? In some cases, I may have to be more than 100m distance from stage.

100 Meters is the GbE (Gigabit Ethernet) maximum distance specification. For longer distances we recommend either inserting Gigabit Ethernet switches at regular distances less than 100 meters, using them as repeaters or better Copper to Optical (Ethernet to Fiber) adapters at each end (can be used with optical fibers up to roughly 600 meters or 2000 feet on a single hop.). A good reference starting point where this is all explained briefly is:
Customers should refer to the specs of the manufacturers of the CAT5e or CAT6 cables that they deployed or plan to use, as Merging only support those cables.

Merging has certified the TP-Link MC200 (NET-TPL-MC200 on pricelist) for Gigabit Multi-Mode Media Converter, and the TP-Link MC210 (NET-TPL-MC210 on pricelist) for Gigabit Single-Mode Media Converter.

Is it possible to use one HORUS/HAPI for a few DAWs inside the RAVENNA network?

Yes, RAVENNA allows sharing of common sources (inputs) between all devices on a Network. Outputs cannot be shared.

Can RAVENNA and OASIS run on the same physical network?

No, we recommend to use a dedicated switch for Audio over RAVENNA, separate from any other network.

I/O Configurations and Features

I saw two headphone jacks at the front of the HORUS/HAPI. Are those of the same sonic quality as the converters on the DA-module card?

The two headphone Jacks use a different D/A, so they are not exactly of the same level of quality as the Line Out themselves but still quite decent.

Why there are such different impedance specs between Hapi MKII and Anubis?

Anubis and Horus/Hapi have two very different Headphone output topologies:

Anubis is designed to provide two individual and totally independent stereo outputs (with different programs and different levels), which are called Headphones 1 and Headphones 2. While Horus and Hapi have also two outputs (6.35mm and 3.5mm stereo jacks), these outputs share a single Headphone Amplifier. It is therefore adequate to decouple each output from each other by a reasonably sized source resistance, so that in the event one was to connect two headphones at the same time to those jacks, there is no mutual influence of one Headphone on the other one. With near zero Ohm output impedance (such as designed in the Anubis) the above relative mutual immunity would be impossible.

Hapi MkII benefits from a significantly lower output resistance (20 Ohms) than its predecessor Hapi (75 Ohms).

Can the headphone jacks be used without installing any DA-module card?

Yes, at all PCM sampling rates, including DXD.
For Horus - Hapi :
In DSD64 and DSD128 modes, the headphones works, but the gain slider for level adjustment is not supported.
The headphone jack does not support DSD256. 
For Hapi MKII :
The headphone works up to DSD256 (including the gain slider).

It is possible to divide the HORUS/HAPI channels between different DAWs?

And control them separately?

With respect to control, the answer is also yes when controlling different portions (channel groups) of a HORUS/HAPI. However when several controlling devices (web pages, control surfaces, etc.) want to address the same channels, some rules will have to be established in terms of access priority and authorization as is typical with any properly managed Network architecture.

For example, can a 24 AD/DA version have the 24ADs being recorded on two computers/DAW’s at the same time and in addition route them directly out to the MADI-stream as well simultaneously (To FOH)?

… And at the same time record 64-channels from the MADI-stream (FOH) to the computers/DAWs 
YES! Although the possible track count being recorded will depend on the DAW’s ability to record lots of tracks. (We know that Pyramix can do it though)
… While sending to 12 channels HORUS/HAPI D/A from computer 1 and 12 D/A from computer 2? 
Yes but note that the HORUS/HAPI modules can be routed (at the moment) in blocks of 8 channels. So you could route 8 channels from DAW1 and 16 from DAW 2 etc.
Please have a look at our Advanced settings guide. 

So, the AD/DAs and other I/O modules need to be configured in groups of 8? Or can you route individual channels (Like common talkback/listen)?

We have decided for the moment to simplify the routing capabilities (Analog - AES - MADI - RAVENNA) by groups of 8 consecutive channels, except for the Headphone monitoring that is just a stereo signal.

How are Gain, Phase, Phantom Power etc. controllable via RAVENNA from within Pyramix?

Since Pyramix v8 the gain and other parameters can be controlled from within the Pyramix mixer.
See this page for further details about Pyramix and Horus/Hapi firmware versions and compatibilities.

Can you route, for example AES and AD/DA (I/O) straight to MADI via DSP while they at the same time go to RAVENNA ?

Yes, Any input can go to any combination of outputs.
How Can I set it all? Do I have to do it on the front panel of the HORUS/HAPI unit?

No, Each HORUS/HAPI unit can be configured and controlled by web page access which can be viewed and manipulated using almost any standard web browser (Chrome recommended/Internet Explorer does NOT work)

Latency: It’s to my understanding that the HORUS/HAPI has sub-millisecond latency, but do you have an exact number?

I'm afraid we may be confusing several functionalities here. In terms of A/D to MADI or MADI to D/A converter, HORUS/HAPI has just a few samples of latency (well under a millisecond) so that should not be a problem.
See Horus or Hapi User manual or Hapi MKII User manual - Modules latencies section for the exact latency values.
The main issue about latency is actually not with the converter itself but with the driver to/from the application. So the latencies noticed will vary greatly between the MassCore and the ASIO based usage of the HORUS/HAPI.

So then, what is the Latency of a HORUS/HAPI being used with RAVENNA and MassCore?

The Latency from live in to live out of a HORUS/HAPI being used with our MassCore engine can go as low as 1.33ms (currently at 5.3ms for standard latency running mode)

What is the latency of the HORUS/HAPI if used with the RAVENNA ASIO/CoreAudio Driver?

As ASIO and CoreAudio are “host-based” processes, the latency when using the HORUS/HAPI with these drivers will increase substantially. This will be similar for any ASIO / CoreAudio solution as it is essentially due to the non-deterministic behavior of the application running under Windows or MAC. 
… And what If simply wanting to use it as an AD/DA and take the converted analogue signal out of the MADI or AES outputs?

Then we go back to the original statement, “HORUS/HAPI has just a few samples of latency (so well under a millisecond)”.

What is the maximum channel count for RAVENNA/HORUS/HAPI @1fs? How about at DSD sample rates?

We've deployed RAVENNA under MassCore with an equivalent capacity of 384 x 24-bit channels equivalent at 1 FS (ie 44.1 or 48 kHz) and this bi-directionally. That corresponds to 48 channels at DXD or 384 kHz rates. Refer to our recommended RAVENNA configurations.  Detailed Computer installation guides are also available for the validated configurations, contact

Can the RAVENNA ASIO driver work on any Gigabit port like the built in of a Mac Pro or PC?

Yes, it is only our own MassCore implementation that will require a dedicated Gigabit adapter card for the best performance.
A dedicated network card is still recommended for better performances. 

What are the Specifications for the ASIO/CoreAudio Driver? What can I achieve with it?

This ASIO driver is expected to allow up to 128 bidirectional I/O channels of 24-bit (at 1FS). It will of course not allow lowest latency operation as with MassCore implementation but for many applications, this will be fine.  The number of channels will also depend on the system performances.

Can I use the ASIO or CoreAudio Driver with DAW Applications other than Merging’s Pyramix and Ovation? Like for instance AVID Protools or Steinberg Nuendo etc?

Yes! These drivers were specifically designed to make sure that regardless of the tools you use you will be able to incorporate HORUS/HAPI into your workflow!

Does the ASIO driver allow a DAW to sit on a RAVENNA network as a standard node? I mean, if I had 2 DAWs using the ASIO driver, could I connect them both to one HORUS/HAPI to do parallel recording?

Yes, you can record a HORUS/HAPI stream from several ASIO driver.


What about the HORUS fan, surely that means the box makes too much ambient sound ?

There will be a fan, capable of turning itself on gradually whenever the internally sensed temperature exceeds pre-set limits. We do not want our customers to have their HORUS audio characteristics degrade or even fail above certain temperature!

Fortunately, Merging has spent enormous time carefully engineering each and every component of HORUS/HAPI to be drawing as little power as technologically possible these days, without compromising the sound at any time.
We therefore expect our users to enjoy their HORUS as an essentially "fan-less" device since the fan will in many (if not most circumstances) never switch on. So the fact that we still planned a fan is to be more seen as an "emergency" fan that :

  • will only turn on when absolutely required (for instance in situations of high ambient temperature, with many phantom-powered connected microphones drawing a lot of aggregate power)
  • even when turned on, we are engineering it in such a way that the fan will be essentially inaudible.

Designing a 2 U unit which offers, in such a compact space, up to 4 MADI ports, 24 channels of AES I/O and 24 channels of MIC preamplifiers along with ultra high-quality A/D and up to 24 channel of D/A conversion, not even mentioning RAVENNA, a backlit TFT touchscreen and an optional Redundant Power supply with no decent "heat-control" would just have been ludicrous. As a matter of fact, we expect a fully loaded HORUS to still not consume more than 50-Watt to maximum of 60-Watt total.

In a configuration where HORUS/HAPI is a stagebox connected to a PC running Pyramix, can I use Pyramix native or do I need a Mykerinos card?

Starting with Pyramix V8, there is no need for a Mykerinos card.
RAVENNA can be connected either :

  • Via our own PCIe x1 Network RAVENNA/MassCore accelerator card NET-MSC-GBEX1 for Pyramix MassCore operation (available in Pyramix V8 Beta today). Allows a maximum of channels with lowest latency
  • Standard computer GbE NIC in ASIO or Core Audio for Pyramix Native and/or other third-party applications (expected ot become available over the summer). Will probably allow up to 64 channels at 1 FS (32 channels at 2FS, 16 channels at 4FS) with moderate latencies, as usual with existing Native audio applications.

Does HORUS/HAPI have a DSP mixer or any other way of controlling/setting up low-latency monitoring?

No, HORUS/HAPI provides a low latency router, which may be used for direct monitoring in certain circumstances.

How does HORUS/HAPI process DSD to and from Pyramix over RAVENNA?

Internally, the HORUS/HAPI actually processes and moves the DSD signal in a 352.8kHz / 32bit stream. It is not converted to actual PCM, rather the raw DSD data is just transported in this manner.

  • A DSD64 stream actually only uses 8 bits of the possible 32bits.
  • A DSD128 stream uses 16bits
  • A DSD256 stream uses all 32bits.
  • The DXD (352.8kHz PCM stream) uses the 24bits

All of these DSD stream types are actually multiples of 352.8, itself being a multiple of 44.1 kHz.
Therefore, for the HORUS/HAPI mainboard there’s no real difference between all these 4 formats. That’s only the AD and DA that are configured automatically and transparently according to the streams they are receiving audio from or sending audio to.
In fact as far as only playback or DA is concerned, you can set your HORUS/HAPI unit to DSD or DXD, that’ll make no difference. It’s only the way the Pyramix projects are configured that will define if audio is sent directly as DSD (when using DSD projects) or converted to DXD (when using DXD projects)

Are Horus and Hapi DoP capable ?

DoP means DSD Over PCM; the goal of this "standard" is to carry DSD signal over regular PCM link; 
Horus/Hapi can carry DoP over all of its digital links (AES, MADI, …), as the device is transparent.
But, as Horus/Hapi doesn't know what DoP is, it cannot decode it before sending it to the DAC.
In summary, Horus/Hapi is not a DoP DAC (nor DoP ADC) but it can carry DoP over its digital interfaces.

The computer driver is also implied to answer such question:

  • VAD (macOS): CoreAudio doesn't support natively DSD, so our VAD driver has to decode the DoP to produce "pure" DSD which is send to Hapi/Horus over RAVENNA. 
    Hapi/Horus receives the DSD signal and send it to DAC.
  • ASIO/ALSA(Linux): these two standards are natively supporting DSD format and so they send "pure" DSD to Horus/Hapi over RAVENNA. 
    Hapi/Horus receives the DSD signal and send it to DAC.
    If the customer application, send DoP to ASIO then the Hapi/Horus cannot decode it but it can send it to its digital interfaces.

Note: depending on the flavor of DSD (64, 128, 256, 512), DoP can use several combinations of sample rate and number of channels (


The AES ports connect the AES-EBU I/O using DB25 D-SUB connections (Tascam Digital). This means that AES cables that came with your Mykerinos cables are not compatible with the HORUS/HAPI. You can order those cables directly from our sales partners.