ninbura

What you're really looking for is a mixer. But there's only one line of mixers that accepts USB microphones, that being the Rode Rodecaster line of mixers. They're very expensive, and only accept Rode USB microphones. If it were me, I'd switch to an XLR microphone and either purchase the Yamaha AG03MK2, or the Rodecaster Duo. With the AG03 you'd connect the main USB connection to your laptop dock and set the on-mixer switch to "dry 1/2". This will isolate the mic signal on the mixer and send it to your laptop over the USB connection. Note that with this mixer your Mic will only be sent to your laptop. You will also get system audio from the USB connection to the mixer. you'd need an HDMI audio extractor, as @thrasher_565 mentioned, to connect your video game consoles to the mixer. I'd also suggest running your two consoles through and HDMI switch to you can feed them both into the same extractor, and then into the mixer with a 3.5 to Dual 6.35mm cable. With the Rodecaster Duo, you'd connect USB 1 to your laptop dock, this would give you two inputs and outputs to your laptop (mic in, system & chat audio out). Then you could grab audio from your consoles via the USB 2 connection. I'd connect said USB connection to a USB switch for easily switching the connection between your PS5 & Nintendo Switch. The nice thing about this mixer vs the AG06 is that you can send your mic over the USB connection to your PS5 & Nintendo Switch. However, the mixer is significantly more expensive. Though, it's also of a much higher quality and significantly more feature-rich. Like giving you Mic FX (compressor, noise gate, eq, etc), Bluetooth connectivity, a separate fader for voice chat / game chat, and much more. If neither mixer is within your price range you could try to find something lower end, but I wouldn't recommend it.

You have a few things to consider Nvidia's consumer-grade GPUs have an artificial 5 encode session limit and your driver will need to be patched to bypass that if you're encoding more than 5 unique streams concurrently. The P5000 has dual encoders, you could run into encoder overload depending on your settings when running 8 concurrent 1080p streams on an Nvidia gpu that has a single encoder. Unless when you say "6-8 streams" you're referring to the sources in OBS, which renders encode session limits & encoder bandwidth irrelevant to the conversation. Use Nvidia's Video Encode and Decode GPU Support Matrix for reference as to which GPUs have dual encoders.That being said... the M2000 only has 1 encoder, so if you're currently pulling off 8 concurrent streams with that you should be good. Unless you want to increase quality via presets or need more headroom in the future. I'd consider using FFmpeg for your streams instead of OBS if you aren't compositing sources. FFmpeg is superior to OBS if you're directly streaming a source to a platform because it only hits the encoder on your GPU & not the 3D engine. YouTube supports AV1 ingest, so with a 40-Series card you could greatly reduce your data rate as AV1 is much more efficient. Also, the 4070 TI and everything above it in the 40-Series has dual encoders, which is relevant if #2 is actually a problem for you.

Worth noting that with some ingenuity you could probably get a 4K setup working, like by sharing the same encode session for streaming and recording. But if there's any reason you're encoding more than 2 streams of 4K60 I would recommenced getting one of the cards with dual encoders.

This also applies to the RTX 4070 desktop chip, which will be a tempting GPU for a streaming PC due to its price point. If your target stream resolution is 4K+, go with the 4070 TI.

Nvidia just announced their RTX 4070, and with it its specs on their official site. Most notably, for the first time since 10-Series; like-series GPUs don't always have the same number of encoders. The RTX 4070 and below (laptop chips) have a single 8th generation encoder, while the RTX 4070 TI and up have 2 8th generation encoders: For most people this shouldn't matter, but if you're trying to encode more than two 4K streams or more than one 8K stream in real-time the RTX 4070 and below won't be able to handle it. For example, if you're streaming 4K to YouTube, recording 4K locally, and stream 4k to Discord (AV1) simultaneously, you'll overload the encoder. This may also be relevant for Plex users who are streaming high resolution content to multiple clients. To be clear this doesn't affect the quality of your encodes, you simply have less encoding bandwidth to play with. They haven't updated it to show the RTX 4070 yet, but Nvidia's Video Encode and Decode GPU Support Matrix is useful when comparing GPUs encoding / decoding potential.

Yeah, despite the extra port it sounds like 40 series still only supports 4 heads, see answer.

That's what I'm currently doing per the original post.

Yeah AMD is better in that regard, I use all the Nvidia technologies though NVENC, RTX Voice, DLSS, etc.

This isn't true unfortunately, modern Quadros are also limited to 4 displays. I was looking into this earlier this year.

Hey there, if anyone out there has their hands on a Tuf / Strix 4090 I would much appreciate if all 5 display outputs were tested simultaneously I run dual GPUs in multiple computers to get a reliable 5th display, wondering if it's finally possible to do 5 displays with a modern Nvidia GPU. My guess is that you'll still be limited to 4 heads despite the extra HDMI port, but a man can dream.

RME's Babyface Pro is cheaper than their ADI-2 and has DSP with EQ. It's only a 3-band EQ per-channel, but with a loopback cable on the ADAT port you can apply an EQ to the input and output giving you 9 bands. Technically if you keep looping the signal back you can have up to 27 bands by the time it comes out of the phones output. The delay on physical loopback with RME interfaces is so low that it's imperceivable. The Babyface Pro is also more capable in terms of routing, IO, when comparted to the ADI-2. Of course the ADI-2 has the highest end phone output, but for OP's headphones of choice the output on the Babyface should be plenty. Almost all my audio gear is RME (MADI FX, UFX+, RayDAT, ADI-192 DD, ADI-648) and I highly recommend it. Didn't even think to suggest it as price for their cheapest interface with DSP is outrageously expensive for most people, but I guess that's what you're saying.

Yeah a lot of audio gear is built to be rack mounted, I'd throw something like this under the desk.

Hardware equalizers can be had, but even the lower end ones are fairly pricey: https://www.amazon.com/gp/product/B004Q9DANA/ref=as_li_tl?ie=UTF8&camp=1789&creative=390957&creativeASIN=B004NDHZTY&linkCode=as2&tag=ledgernote-20&linkId=KUVUVIRKLQAGUWWC&th=1

This is actually a receiver; not a transmitter, it's a decent one though They do sell an "upgraded" transmitter, but as far as I can tell it's just that there's a small screen now, it's $20 more though. Haven't tested it myself but it feels like it's probably not worth unless you're really utilizing the dual headphone capability as it seems to give you better visibility and seemingly independent volume control.

I use one of these on the outset of my audio rack: https://www.amazon.com/gp/product/B0838YPSZT/ref=ppx_yo_dt_b_asin_title_o06_s00?ie=UTF8&psc=1 Bluetooth transmitters like these are sometimes misconstrued to have low quality, but that's because the analog to digital converter in them is usually gated and generally low quality. If you come in through the optical port signal integrity is perfectly maintained, the audio isn't converted from digital to analog until it hits the headphones. Most PCs have an optical output, super easy to setup with the included cables, should be able to power it off a USB port on the PC too. Supports all the best Bluetooth codecs and will have much better range than a tiny little dongle. It should be noted that this will not support two-way communication, so you can't use the mic on said Bluetooth device if it has one. But it should also be noted that quality is greatly degraded when using two-way communication on any Bluetooth headset. Both the input and the output are converted to mono signals with extremely low bit rates. If you're attempting to use a Bluetooth headset with Discord on PC or something I would highly recommend using an external mic separate from the headset. You could still use a wireless one if desired, like Rode's Wireless GO II system, or the Mod Mic Wireless.

IMO the "best wireless solution" is the SteelSeries Arctis Pro Wireless if we're talking about the consumer and not professional space. It has very low latency in relation to its price point at 16ms, this is much lower latency that even the fastest Bluetooth protocols. To get lower latency than this you need to spend sometimes thousands of dollars on professional wireless audio systems, like the Shure PSM 300 system, or the Letrosonics Duet M2. What separates the Arctis from other wireless solutions is: The dock allows for USB and Optical inputs simultaneously. Meaning you can plug your PC into it twice, once via USB, and another time via optical. This would allow you to send voice chat and game volume as separate which can then be mixed for volume via the dock. This also allows you to connect an external mixer to the dock while still utilizing the microphone and wireless capabilities. For example, you could send the L/R outputs of a mixer like the Yamaha AG06 to an analog to digital converter, and then into the dock. At this point you could hear everything coming from the mixer in the headset wirelessly. You could have multiple PCs, consoles, your phone, and other sources connected to said mixer. You can direct monitor the mic with zero latency, so you don't feel like you need to scream over whatever you're listening to. Not to say that other headsets don't have direct monitoring as a feature, just not all do. The headset has two batteries and can be hot swapped on the fly. While you're using the headset another battery can be charged in the dock, when the one in the headset gets low it takes seconds to swap out. If some of these more nuanced features don't matter to you there are cheaper alternatives of course, but when it comes to the best of the best, I think this is it really.

I do not recommend this setup anymore, everything worked in theory but in practice noise was a big issue. I've now worked with over a dozen people on multi-PC audio routing setups and there's 3 proper solutions: Using the Shure MV7 with a Yamaha AG06 (or comparable mixer). This is by far the cheapest solution to properly routing a single mic to two PCs, since the MV7 has a USB and XLR connection that can run simultaneously. Using a mixer with sub groups / sends, compounded with a balanced line level splitter (Whirlwind or Radial Engineering), and balanced line level interfaces (Scarlett 2i2, Behringer UMC202HD). This is much more expensive than the previous solution and can still result in noise without proper cabling and routing. While it does allow you to use your mic of choice, I almost never recommend this solution. The routing is highly complicated, and completely decoupling monitor and record volume for your mic isn't possible. Using RME interfaces on both computers interfaced with ADAT optical connections. This is by far the most expensive solution, but also the most expandable and highest quality. This allows you to use your choice of mic like the last solution, but without the risk of noise as the connection between the two interfaces is carried digitally. Send me a direct message if you want more info.

Never said it was great reasoning, but I do believe that is their front.

Yeah the only reason I would have hope that a patch would work at all and persistently is because that has been the case for NVENC: https://github.com/keylase/nvidia-patch/tree/master/win Though funny enough the last few drivers have broke the patch and the main dev is MIA: https://github.com/keylase/nvidia-patch/issues/440 So yeah if I want to be on the latest drivers this is probably problematic in the future, not really an issue on my capture PC where the drivers have been stable for eons regarding NVENC. Seeing as I don't want to take a step down from my 3090 the only real solution is probably running a GT1030. One of the displays I use is just for Chrome dev tools and file explorer while I'm developing, could probably handle that fine without too much oddity. I also have separate boots for gaming and programming so I could just leave the GT1030 disabled on the game boot. All annoying nonetheless, though I think NVIDIA's reasoning with both the NVENC and this screen limitation is to dissuade professionals / business from eating up consumer cards instead of buying Quadros, which largely funds their operation / R&D for Geforce cards.

Could be that old GPUs didn't have the limit or something, but there's definitely one on my RTX 3090.

Try connecting a 5th monitor.

I'll look into patching it, and I'll contact NVIDIA and bitch to them (obviously this wont do anything lol). They did officially raise the NVENC limit from 2 to 3, maybe it's time to raise the monitor limit. May just have to get a passive GT 1030 or something, but almost every time I've tried running dual GPUs I run into other issues

Wait... doesn't that conflict with this other information? From my understand daisy chaining uses MST technology, I've got 5 monitors, one connected to the first DP, one connected to the HMDI, 1 connected to the 3rd DP, and then 2 connected to the final DP via an MST hub: https://www.amazon.com/gp/product/B095H9BN24/ref=ppx_yo_dt_b_asin_title_o02_s00?ie=UTF8&psc=1 Resolutions: 4K60, 4K60, 3440x1440@100Hz, and then on the MST hub I'm trying to do 2x 1080p monitors.

Yeah I can understand why you'd want to switch because of that. There wasn't a patch for NVENC a few years ago either, but I bet it's a similar situation. I'm actually trying to use this 5th display as just a "fake" display run by and EDID emulator, need somewhere where I can full screen my OBS preview so that I can share it in applications like Microsoft Teams and Discord. Maybe there is a software solution to this that doesn't involve running 5 displays on the 3090.

Frustrating, maybe I'll look into how NVENC is patched and see if there is any way around this limitation too. Thanks for the info.