How is your setup? FF800 connected to PC via Firewire? What Firewire card is in use?

Could you please post a screenshot of the driver settings and make a video of how the noise sounds?

Maybe it's about time to sell the FF800 and to replace it by an UCX II or UFX II.

Currently, you can still get a nice amount of money for the used device as you can get Firewire still working.
On the other hand, Firewire is a dead horse. So, I would sell as long as you can get some money when selling it.

By this, you will get some benefits of a new recording interface with up-to-date converters, mic gain of 75dB, DURec, Autoset, etc.

Quick comment back

> Heh you can only connect FF via FW right ;-)

I wanted to be sure to understand your setup. Some people do not tell information about their whole setup, maybe you run it in standalone mode…

> If I heard an AB that blew me away...sure...but havent had that to date..

Just by the way because you mentioned that you see no reason to change. If money is not an obstacle, then I would always recommend today to a combination of RME recording interface (UFX II / UFX+) and ADI-2 Pro converter for monitoring. KaiS reported that he was very satisfied, especially with the D/A filter setting “Slow”: https://forum.rme-audio.de/viewtopic.ph … 8#p178258.

Here is an Excel list that I created to list all important information about RME USB/FW/TB interfaces of the past and the current models. There you see also a line where you can read the converter latency of all products and other useful information.
The blog article: https://www.tonstudio-forum.de/blog/Ent … B-MADIfac/
The direct link to the Excel file: https://www.tonstudio-forum.de/attachme … 4-08-xlsx/

This kills two birds with one stone. Excellent converter quality and speed at the main interface. USB2/3 is the de-facto standard today and is fully supported by Windows. However, you should follow the chipset recommendations for USB3 (UFX+).

It is generally a good idea to install a separate USB3 card for the recording interface only in certain cases. The Sonnet USB3 card with FL1100 chipset is recommended. Clean board layout and well-functioning drivers that are integrated directly into Windows 10 using the more efficient interrupt handling scheme MSI (message signalled interrupts). This separates the audio interface from the rest of the USB2/3 infrastructure of the motherboard and prevents problems that can arise by connecting additional USB devices such as Bluetooth adapters.

You can also have a look at the converter shootout, where the ADI-2 Pro compares very well with 4x more expensive AD converters (and they can only do AD) and may even be superior to them in some use cases. The original WAV files are better viewed via the links below the video: https://www.youtube.com/watch?v=doHG32aXBDY

I am using two ADI-2 Pro R BE in my setup. One in the recording corner and one in front of my High-End HiFi. Even the 1st ADI-2 Pro years ago had a slightly better presentation compared to Accuphase own DAC module for the E-600 Class integrated amp with B&W 803D3 as speaker.

I need for example quick A/D and D/A conversion. Be it for “near-realtime” monitoring or to build a parallel effect loop for my Guitar amp. For this purpose, I am using an UFX+ to read from preamp (effect-send) and send audio directly to power amp (effect return), for this I require the converter to be as fast as possible. Then I add 100% wet FX to the port leading to effect send (power amp of the Marshall combo). This results in a much better sound instead of noodling the whole preamp signal though the effect processor before it enters the power amp of the Marshall combo amp.

If you compare converter latency of FF800, then you can see that it changed over the years significantly, its less than a third (27%) of the FF800 as you can see here (values taken from manual chapter "Latency and Monitoring":
FF800 @44.1 kHz: 44 samples A/D, 28 samples D/A => 0,98 + 0.63ms = 1,61 ms
UFX II @44.1 kHz: 13 samples A/D,   7 samples D/A => 0,28 + 0.16ms = 0,44 ms (27%, less than a third)
FF800 @96.0 kHz: 44 samples A/D, 28 samples D/A => 0,45 + 0.29ms = 0,74 ms
UFX II @96.0 kHz: 13 samples A/D,   7 samples D/A => 0,13 + 0.07ms = 0,2 ms (27%, less than a third)

Back to the concrete issue

Regarding your issue and clock settings, it's ok, the FF800 is clock master.
It indeed sounds as if the interface is not able anymore to properly process audio at this point.

Was this initially a 96 kHz recording? What happens
— if you record in single speed
— if you restrict the number of channels to transfer only analog ports over Firewire.

To find the root cause for this can be time-consuming.

You mentioned that you changed the system, so you have another mainboard, BIOS, Chipset, drivers, GPU.

Did you optimize your PC for audio?
Did you perform LatencyMon measuring to identify latency problems caused by bad drivers?

The latter can block a CPU core and if an audio process (driver, application) is being scheduled to run on this CPU core, it can create issues if audio cannot be processed in time.

Another issue was for me (with my system) that LatencyMon results were good, but I got audio glitches for around half a second. It took me 2 years to find the root cause. It was the nVidia graphic card. The cards also have an energy saving of their own, but nVidia doesn't deliver official tools anymore to turn energy saving off on their graphic cards.

I needed a special tool from a Russian developer to deactivate it.

A couple of questions regarding your system:

Did you turn off energy settings in the BIOS, C-States (sometimes also P-/T- States)?
Did you turn off clock spread spectrum in the BIOS to turn off little clock variations?
Do you use Turbo mode, this sometimes brings 100-200 MHz higher clock?
Is the Clock in your PC stable, or does it change?

What energy profile do you use in Windows 10, you can try to activate Ultimate Performance, as Admin you can make it visible/available by using this in the dos box (cmd):  powercfg -duplicatescheme e9a42b02-d5df-448d-aa00-03f14749eb61

A collection of threads about performance troubleshooting here
https://forum.rme-audio.de/viewtopic.ph … 04#p186404

In some cases, it can help to move the Firewire PCIe card to a different PCIe slot.
The “better” PCIe slots are those with PCIe lanes coming directly from the CPU (not chipset) and where the PCIe lanes are not share by other components. It depends on CPU type and mainboard layout. The mainboard manual should contain a block diagram of the I/O ports, how all components are connected to each other.

You should also check which Firewire chip you use. Those from TI (Texas Instruments) are the best, but even they developed a product which had issues, did not react to interrupts in time. Unfortunately, many PCI/PCIe card vendors used the XIO2200A for their designs. A good card that I can recommend you is the one from Exsys using the XIO2213B chips.
https://www.exsys-shop.de/shopware/en/c … le-bracket
https://geizhals.de/exsys-ex-16415-a375938.html

Other possibility, PSU issues on FF800

In other threads, I read that some noise could theoretically also come from the PSU. A PSU replacement should be not too expensive so that you can at least exclude issues coming from there.

Regarding your comments about devices / sound

If you tried already so many components, then one or too more should not be too hard. I would be very curious what you would say regarding UFX II or even UFX+. MADI has some advantages over ADAT in terms of scalability and placement of additional Mic Pres or other components. To have USB3 and Thunderbolt can also be an advantage, then you have more flexibility in terms of connection options towards the PC.

Live recording & RME DURec (Direct USB recording)

I think this could be extremely useful to you. If you perform live recordings, then I think a recording interface with DURec is a must. You can make a backup recording of all channels on a USB disk or stick. If you should have a hand situation or bad audio, maybe caused by Firewire or USB connection or issues on the PC (which is likely in this case).
Because then you still have the recording on the stick in safe harbor, as this audio flow stays local to the recording device. You can even use the recording interface in stand-alone mode without a PC. Select from 6 config profile on the device and make adjustments to the setup.

Regarding sound quality

I know many people who are impressed of the sound quality of UFX II/+ and ADI-2 Pro.
If I were you, then I would really give the combination of UFX II or + and ADI-2 Pro FS R BE a try. What can you loose?
Nothing. I think you won't be disappointed. Besides my high end HiFi I am using good near-field monitors, the Geithain RL906D which are ideal for the near-filed as its position is very uncritical. It's often being used in outside broadcast vans, where you in some cases only can provide parallel positioning at the wall. The converter quality of the last recent products is extremely well. Transparent, not adding any unwanted mojo. Good/stable bass fundament.
The Mic Preamps are superb, UFX II/+ and 12Mic use the same design and the 12Mic got excellent reviews.

FF800 is from 2004. It is certainly within the realm of possibility that something has changed in the past 18 years, after all.

> I installed Latmon, says no issues, adjusted bios...issue still there

Nevertheless, the values must be interpreted correctly. Bad / too high DPC values are an indication that something is not working optimally, so you don't have enough headroom under load. You can see some charts for comparison in some threads (the link I recommended regarding PC performance troubleshooting).

I'm not trying to persuade you, just making some suggestions that might be of interest to you, the final decision is of course up to you.

Over the years I got a good overview of the whole RME product range, and I am skilled in combining the products cleverly.

> Which is the actual link...there are so many there?
Yes, every thread differs a little and I have not the time to write everything again and again.
Important for the troubleshooting is that you know how a computer works.
Usually nobody does this and thus is a little lost with such tools and interpreting the results.
So, here and there, I try to give some background information so that everything becomes clearer.

Regarding your results.

It looks fine, but running it for 35 seconds is too low and do measuring on an IDLE system.
Because LatencyMon already simulates a DAW load.
So start no other application and perform no mouse pointer movement.

If you know your system and only want to perform quick checking whether it makes big jumps in terms of results, then 5 minutes is ok. Otherwise, I would recommend letting it run for at least 10 min if things needs to go fast. Otherwise, you need at min 1-2 hours to get scheduled jobs in the hope that the jobs start at lease once in an hour.