Question / Help Q: 2 Monitors and 1 Capture Card - Refresh Rates and Setup

[VyrilGaming]

Hello all,

Ok so I'm planning on setting up my stream like the following,


Main Monitor - 144hz
Secondary Monitor - 60hz
Capture Card - 60hz

GTX 970
i7-4790k
Avermedia Live Gamer HD

Will I be able to play on my 144hz monitor while having a second monitor with an extended screen, and the capture card capturing my game and / or second screen?

I'm at work and wondering if this is something I can accomplish with 1 video card. Also what do you think the performance hit will be on this?

 

[dping]

Hello all,

Ok so I'm planning on setting up my stream like the following,


Main Monitor - 144hz
Secondary Monitor - 60hz
Capture Card - 60hz

GTX 970
i7-4790k
Avermedia Live Gamer HD

Will I be able to play on my 144hz monitor while having a second monitor with an extended screen, and the capture card capturing my game and / or second screen?

I'm at work and wondering if this is something I can accomplish with 1 video card. Also what do you think the performance hit will be on this?

capture cards are not for single PC setups. your system will be more loaded down by doing this. capture cards do what they say, they capture.

this is ideal when capturing from a gaming PC to a streaming/recording PC.

 

[VyrilGaming]

capture cards are not for single PC setups. your system will be more loaded down by doing this. capture cards do what they say, they capture.

this is ideal when capturing from a gaming PC to a streaming/recording PC.

Actually you're wrong, there is such thing as pass through. Which I use to stream with and it increased performance in my Single PC setup. Instead of having the Single CPU capture and Encode the capture was done by the capture card and encoding by the CPU which reduced the overhead on the CPU.

This was a setup I had on my i5 which didnt have the HT and I had an older video card. It performed decently but the i5 just couldn't encode good enough.

Offloading the capturing to a capture card allows you to have better compression on your stream and overall quality. I know the Live Gamer HD only does 30fps @1080p but I don't care about that at this moment.

I'm just wondering if anyone has had luck with my above scenario. So I don't have to use pass-through and still get my 144hz refresh rate on my gaming monitor but capture at 30fps.

 

[FerretBomb]

Actually, you're wrong.
Not with OBS. A capture device in a single-PC setup will cause more overhead and worse performance as compared to a Game or Window capture.

OBS Game Capture grabs directly from VRAM and copies to another point in VRAM (OBS uses video memory for scaling/compositing the final frame).

A capture device has to take the video the card puts out, re-capture it, transfer it across DirectShow to system memory, and then OBS copies it into VRAM for scaling/compositing, right back onto the video card it just came from, but doing a HUGE and unnecessary loop through the rest of the computer for zero reason. Nothing is offloaded, overhead is increased.

A capture card used with OBS is worthless unless you plan to have a 2PC setup, or stream console games.
The on-card video encoders that Avermedia/Elgato tout only work with their software, NOT OBS.

No, unless your capture card supports 144hz (and almost none do; certainly not Avermedia junk) the passthrough will force-negotiate it down to a 60hz signal.
Some video cards (mostly nVidia) can clone a given monitor to a second output at a different refresh rate (allowing output 1 to your 144hz monitor at 144hz, and output 2 to your cap card at 60hz). So if you have an nVidia card and aren't monitor-count capped, that could be an option.

It's a semi-common question if you used the Search function.
Short version, unless you want to play console games, you wasted money on the capture card. And fell for a line of marketing BS.

 

[Boildown]

1) Actually you're wrong, offloading to the capture card isn't supported by OBS. Or by anything other than the native Avermedia software when using Avermedia capture cards. My understanding is that the code is proprietary to Avermedia and not available for OBS to use, even if the developers of OBS wanted to.

2) Game capture is straight up better than using the passthrough option of your Avermedia capture card in a single PC with OBS. So use game capture, not a capture card. If possible, return it and get your money back.

3) I should point out that the compression (the quality for the bitrate used) by the on-board chip on the Avermedia is fucking awful. If you stream with this using the RecCenter or whatever its called these days, it'll look like complete garbage. Because 2000kbps compressed by x264 is many times better quality than 2000kbps compressed by the p.o.s. onboard chip. If you just want to record to hard drive, then you can crank the bitrate up on the Avermedia and have it be ok. But not with OBS; see 1).

 

[VyrilGaming]

So when I was encoding without a capture card before and the best I could do was very fast, but then added the capture card and was able to encode with medium, and have a noticeable difference in stream quality?

Both scenarios using OBS. Not Avermedia's software.

 

[Boildown]

Post log files from before and after, along with VODs or sample recordings from each. Maybe you're on to something. Or maybe just on something.

 

[dping]

So when I was encoding without a capture card before and the best I could do was very fast, but then added the capture card and was able to encode with medium, and have a noticeable difference in stream quality?

you were probably doing something wrong. I'd really like to see the logfiles from both using the same game or even a benchmark captured. one with your avermedia and the other with a straight game capture.

 

[VyrilGaming]

I'm not against changing my setup. My only first hand knowledge is with the above scenario where I did see performance improvement.

I didn't have a i7 before and I don't want to take a performance hit on the games. I dont have the second PC built for streaming at this time.

 

[VyrilGaming]

you were probably doing something wrong. I'd really like to see the logfiles from both using the same game or even a benchmark captured. one with your avermedia and the other with a straight game capture.

I spent HOURS messing with settings reading and configuring. It was a straight up and clear performance increase with the capture card.

Additionally my games suffered and so did the stream without the capture card.

Unfortunately this was over 2 years ago. So no logs, I may have some VOD. I will look.

 

[FerretBomb]

It's just pretty hard to believe, given that it goes against the underlying logic of how it works in both cases, as well as user testing across the board. It more indicates another underlying problem with your setup, like if you were using Monitor Capture, and not Game Capture. Because MonCap under Win7 performs horribly, so a capture card would probably see a reasonable-to-significant performance increase. But not against a Game Capture.

 

[dping]

I spent HOURS messing with settings reading and configuring. It was a straight up and clear performance increase with the capture card.

Additionally my games suffered and so did the stream without the capture card.

Unfortunately this was over 2 years ago. So no logs, I may have some VOD. I will look.

So you said you didn't have the i7 before...This would be why you weren't able to run med preset prior.

I didn't have a i7 before and I don't want to take a performance hit on the games.

Anyway, If you are going to make statements like

Actually, you're wrong.

then you need to provide some evidence because we here, go by tested and experimenting plus the thousands of logfiles we've seen and issues we've fixed in various setups.

your testing was more than likely with either different hardware or different games or even scenes. I will say I've yet to see a 4790K do medium 720@60fps without in-game performance loss (stuttering or DPC latency issues). most of the time, a 4790K OC'd will do fast preset with 720@60 with some games.

long story short. your capture card is seemingly useless if not putting more strain on your PC. The load has to go somewhere.

 

[VyrilGaming]

So you said you didn't have the i7 before...This would be why you weren't able to run med preset prior.

I didn't have a i7 before and I don't want to take a performance hit on the games.

Anyway, If you are going to make statements like then you need to provide some evidence because we here, go by tested and experimenting plus the thousands of logfiles we've seen and issues we've fixed in various setups.

your testing was more than likely with either different hardware or different games or even scenes. I will say I've yet to see a 4790K do medium 720@60fps without in-game performance loss (stuttering or DPC latency issues). most of the time, a 4790K OC'd will do fast preset with 720@60 with some games.

long story short. your capture card is seemingly useless if not putting more strain on your PC. The load has to go somewhere.

I was running 720@30 in my scenarios.

Now I understand the tech behind the i7 and how it's much more reliable to be a stand alone without a capture card.

The hardware was the same except the CC, and the game was LoL. Without CC I would be running 100% CPU usage, and noticeable game performance problems. Then I changed to the CC and was able to run medium 720@30 with about 85-95% CPU usage but was able to play LoL without lag.

When you say the load has to go somewhere. Well the "load" of capturing the video is no longer on the CPU, it's on the card. Then the CPU is used to play the game and Encode.

 

[VyrilGaming]

All in All, this could all be due to the fact it's beneficial for a mid range PC with no Hyper Threading on the CPU.

I'm not arguing that this could be useless when it comes to a newer setup, but it was factual that in my scenario it improved performance on my i5 2 years ago.

 

[dping]

I was running 720@30 in my scenarios.

Now I understand the tech behind the i7 and how it's much more reliable to be a stand alone without a capture card.

The hardware was the same except the CC, and the game was LoL. Without CC I would be running 100% CPU usage, and noticeable game performance problems. Then I changed to the CC and was able to run medium 720@30 with about 85-95% CPU usage but was able to play LoL without lag.

When you say the load has to go somewhere. Well the "load" of capturing the video is no longer on the CPU, it's on the card. Then the CPU is used to play the game and Encode.

This is an incorrect assumption and is not enabled in OBS for a capture card to encode.

 

[FerretBomb]

When you say the load has to go somewhere. Well the "load" of capturing the video is no longer on the CPU, it's on the card. Then the CPU is used to play the game and Encode.

Nope. Again, all that an OBS native capture does is copy a video frame from one spot to another in VRAM.
OBS can't access any kind of capture card encoders. It absolutely, positively, 100% will NOT offload ANYTHING from the CPU or GPU whatsoever, and will increase CPU overhead to boot. OBS actually handles video capture sources pretty poorly, on top of that.

So yeah, if it made your performance better, there was something else going on with your setup that needed to be fixed.
Because using a capture device on a 1PC setup is like getting in your car and driving to your local dam every time you want a drink of water, instead of just turning on a tap. Oh, and bringing the dam water home to drink it while standing next to your sink. It's really that obtuse.

 

[dping]

All in All, this could all be due to the fact it's beneficial for a mid range PC with no Hyper Threading on the CPU.

I'm not arguing that this could be useless when it comes to a newer setup, but it was factual that in my scenario it improved performance on my i5 2 years ago.

just post your log from the help menu of your last encode.

 

[Boildown]

Claiming that your capture card in a single-PC solution made things better than game capture is an extraordinary claim. Which gives me an excuse to post this:

[removed image]

Until you produce some evidence to back up this claim, I for one am done here. You are of course free to do whatever you want, but seeing how I've given my advice, if it isn't followed, I just point you back to my advice.

 

[VyrilGaming]

Claiming that your capture card in a single-PC solution made things better than game capture is an extraordinary claim. Which gives me an excuse to post this:



Until you produce some evidence to back up this claim, I for one am done here. You are of course free to do whatever you want, but seeing how I've given my advice, if it isn't followed, I just point you back to my advice.

In any case I don't know why you have such a issue, I don't think I've seen somebody get so riled up over being wrong.

I for one don't care if you believe me or not since I don't even use that system any more and my initial question was answered, I have no problem not using a CC since my specs now are able to stream without one.

 

[VyrilGaming]

This is an incorrect assumption and is not enabled in OBS for a capture card to encode.

who said anything about the capture card encoding?

Remind you I wasn't "game capture". That's why I had performance issues. Game Capture just tanked my system unless I was very low resolution.

MEEHH, anyway I don't care anymore. It clearly performed better with it. It still was weak in comparison, but I could play LoL with the Capture Card and couldn't play without. I don't understand how people don't understand that.

I'm not trying to impress anyone with my crappy old rig.