Recording at 70fps ?

[I3reeze]

I'm using Dxtory to record gameplay to disk. I'm currently recording at 1080p 60fps, and Dxtory's overlay shows that the fps being written to disk is more between 45-62 fps. I assume this is because it's difficult to maintain a solid framerate. Would it be possible to record at 70fps in hopes of getting between 65-72 (ish) fps? And would that be a problem for Vegas Pro 13 to read and render at a solid 60fps?

 

[dudeman509]

Is your disk usage pegged at 100% while recording? With it varying that much, I think there is a bottleneck somewhere (CPU or disk).

 

[I3reeze]

Is your disk usage pegged at 100% while recording? With it varying that much, I think there is a bottleneck somewhere (CPU or disk).

Here's my build: http://pcpartpicker.com/user/I3reeze/saved/vwpWGX
My cpu shouldn't be the problem, but the disk may be.

Here are two pics taken of my disk usage when recording, you can also see that the CPU never maxes out.
http://imgur.com/a/sUfIG

 

[dudeman509]

Looks that way. Are you gaming and writing to the same disk? That would definitely max it out as it tries to load new areas and save data at the same time.

Is C a SSD? If so, try writing there as a test and see if you have the same behavior.

 

[I3reeze]

Looks that way. Are you gaming and writing to the same disk? That would definitely max it out as it tries to load new areas and save data at the same time.

Is C a SSD? If so, try writing there as a test and see if you have the same behavior.

Okay so I've been trying to recreate the same results in my previous post all night but I can't seem to get E: to max out anymore.

Here's the results of writing to C: which has the game running on it. Based on Dxtory settings, it's writing at 438 MB/sec: http://imgur.com/BFI5Y6E
Writing to D: which is doing nothing otherwise. Dxtory settings, it's writing at 78 MB/sec (seems low): http://imgur.com/MzJh29O
Writing to E: at 102 MB/sec: http://imgur.com/DGO2DOu

GPU usage is maxed but that may be because I have crossfire disabled. I'm getting 100+ framerates in game though, so I don't think that's an issue.

 

[I3reeze]

I'll re-link my build here: http://pcpartpicker.com/user/I3reeze/saved/vwpWGX

Upon further researching, I've found that using a RAID 0 setup might be very helpful for recording at solid 60fps, and I could probably record in my native 1440p instead of lowering resolution at recording/encoding time to 1080p.

Though, before I move on to other possible solutions, I'd like to hear what you think about my current disk usage.

 

[dudeman509]

I'm trying to figure out the discrepancy between D and E speeds if they're the same drive, yet D is only tasked with writing files. Doesn't make sense to me, unless they're writing at totally different parts of the disk surface (although it shouldn't make THAT big of a difference in % usage).

I do agree that the RAID-0 setup would bring some pretty significant speed gains with it.

 

[I3reeze]

I'm trying to figure out the discrepancy between D and E speeds if they're the same drive, yet D is only tasked with writing files.

I can't write the whole .avi recording to both D: and E: at the same time without setting up RAID 1 but I don't want to make a huge change like that because it will alter any troubleshooting results. Dxtory only allows me to write to one drive at a time unless I write in .rawcap which basically works like a RAID 0 on any available HDDs which I select and then rebuilds the whole .avi file later. I also think this isn't quite what I'm looking for if I want to compare D: and E:

Is there some other test or method which would help you discover a discrepancy?

 

[dudeman509]

I'm trying to figure out the discrepancy between D and E speeds if they're the same drive, yet D is only tasked with writing files.

I can't write the whole .avi recording to both D: and E: at the same time without setting up RAID 1 but I don't want to make a huge change like that because it will alter any troubleshooting results. Dxtory only allows me to write to one drive at a time unless I write in .rawcap which basically works like a RAID 0 on any available HDDs which I select and then rebuilds the whole .avi file later. I also think this isn't quite what I'm looking for if I want to compare D: and E:

Is there some other test or method which would help you discover a discrepancy?

Exact same model of drive, I meant, not tied together. Poor wording on my part. What's the file system between the two look like? Could D be more fragmented than E?

 

[I3reeze]

Could D be more fragmented than E?

D is certainly more fragmented than E due to the folders and programs it contains, but they've both been defragmented with the windows built in program.

 

[dudeman509]

Could D be more fragmented than E?

D is certainly more fragmented than E due to the folders and programs it contains, but they've both been defragmented with the windows built in program.

Games you're recording located on D as well? That would explain the high usage there.

The spikes on the first disk usage graph with E would explain the unsteady frame rates there as well. Do you maintain a steady 60 FPS when recording to C? If so, probably a totally blank E drive, recording to the SSD and then moving it over, or RAID might be your best bet for steady FPS when recording.

 

[I3reeze]

Could D be more fragmented than E?

D is certainly more fragmented than E due to the folders and programs it contains, but they've both been defragmented with the windows built in program.

Games you're recording located on D as well? That would explain the high usage there.

The spikes on the first disk usage graph with E would explain the unsteady frame rates there as well. Do you maintain a steady 60 FPS when recording to C? If so, probably a totally blank E drive, recording to the SSD and then moving it over, or RAID might be your best bet for steady FPS when recording.

This particular game is actually located on my C: (SSD). Dxtory is also located on C: and it is doing the encoding but C: isn't really struggling during the recording process.
I did some testing last night and, in a training area without other players in the lobby, I was able to record at 1440p60fps with pretty much no frame drop. This could be due to a lack of particle effects, but does that say anything about the issues I've been having?