LucShep

This is an easy day one buy for me the Mig-17 is my all-time favorite eastern aircraft. I was hoping also for the AS version, but just the F version is mighty fine. I don't think I've seen it listed here but the F was also capable of using the K-13A missiles (reverse engineered copies of the AIM-9B). Hopefully these will be included as available armament in the module as standard. One request I'll do right away to the devs is, to please bring status updates to these forums, and/or to the official RSR website. Facebook is an utter blight and too many won't ever go there.

Given how sounds are made by different people and teams (ED and 3rd parties), the consistency of in-cockpit sounds varies greatly. Some modules have the engine (or wind and turbulence) sounds drowned inside cockpit, others have them quite exposed (engine, or wind, or turbulence, or all). Some have them perfectly aligned with the "Hear Like In Helmet" audio option (a must for realism, IMO), others seem quite a bit broken by this option (requiring it off). We once had the main and module's sound files (WAVs and OGGs, plus .SDEF) to play around with this separately on our own, but no longer since these all became encripted. Especially in regards to sound volume, such as in-cockpit engine sounds, wind, turbulence and etc, now requiring settings which the user can no longer (re)adjust for custom/personal tastes. There is no way to adjust volume of these different sounds separately. Other than asking for a bit more "in-house QC" (by ED) for consistency purposes, and for all the sounds, I think part of the solution would be to have a few advanced options for Audio, with specific and separate volume sliders, such as: Internal (in-cockpit) - Engine Sounds "0 --- 100" Internal (in-cockpit) - Wind /Turbulence Sounds "0 --- 100" Internal (in-cockpit) - Weapon Sounds "0 --- 100" Internal (in-cockpit) - Avionics/Comms Sounds "0 --- 100" Internal (in-cockpit) - Other Effects "0 --- 100" External - Engine Sounds "0 --- 100" External - Weapon Sounds "0 --- 100" External - Other Effects "0 --- 100"

The new sounds are pretty good but... unfortunately you do need to deselect the "Hear Like In Helmet" option to make them noticed. The problem with this is, in this situation (as with any other module) all the other sounds effects -wind, and etc- are way too loud and exposed, becoming too tiresome and even unpleasant, after a bit. Once you select the option "Hear Like In Helmet" in the options (which is fine in most modules), the F-5E in-cockpit engine sound is completely drowned by all other sounds, like there's no engine sound at all. The sound files are encrypted (otherwise I'd play around it myself). Any way to increase the volume in-cockpit engine sound, please ?

Yes, Cinebench is more focused on CPU performance, you're not getting the clear picture from RAM in it. For RAM, I'd say AIDA64 Memory (paid version) and MaxxMEM2 are better to benchmark that.

Yep! So, you already understand that the 3600 (Mhz) listed there is about the speed of the memory. But the other four main timings also represent important aspects, as you now notice, with the new 3600 17-19-19-39 mem settings. So, in that "17-19-19-39" order.... CL: CAS Latency. The time it takes between a command having been sent to the memory and when it begins to reply to it. It is the time it takes between the processor asking for some data from the memory and then returning it. tRCD: RAS to CAS Delay. The time it takes between the activation of the line (RAS) and the column (CAS) where the data are stored in the matrix. tRP: RAS Precharge. The time it takes between disabling the access to a line of data and the beginning of the access to another line of data. tRAS: Active to Precharge Delay. How long the memory has to wait until the next access to the memory can be initiated. As you can realize, the shorter these timings are, the better. And it's why 3600 17-19-19-39 is a bit better than 3700 18-22-22-42. Ideally, you want the highest speed AND shortest timings, but that does not come cheap (the DDR4 "B-Die" memory prices!) or easy, if you're into RAM overclocking. In reality, it's not like small differences can be perceived (most times they're not) but there can be measurable differences in some aplications, as you now notice in DCS.

BTW, here's a DDR4 OC Guide: https://github.com/integralfx/MemTestHelper/blob/oc-guide/DDR4 OC Guide.md#table-of-contents Lots of interesting stuff, also for sub-timings. If interested, please read it thoroughly, and do not feel tempted to test settings that go beyond the mentioned "safe" settings in it.

OMG.... The info I had was that VCCSA settings were locked for the user. I thought it would automatically re-adjust for XMP profile! I'm am so, so sorry! In this case I need to take back what I said about your motherboard (and any other B660 mobo, that is), because it seems to be doing what it's meant by design. Now it's explained why your process has been difficult! Leaving 1,40v on DRAM voltage, bringing back the frequency to 3600 and tighten the timings is a good idea, but if it actually allows you to use 3700 MHz 18-22-22-42 stable and without any issues (on stock/locked VCCSA) then there's nothing to complain about it. Still, have to say, fully locked VCCSA settings are a very dumb decision by Intel and mobo manufacturers. B660 was supposed to allow for memory overclocking!

I'm actually in the process of modding the NTTR, atempting to give it back the "desatured" (and correct) tones that its initial release (DCS 2.0) had then. I've been on and off with experiments and got nothing to show yet, but it will eventually be done. BTW, this thread explains the current issue with the overly saturated color/tint in the NTTR map of DCS:

hmmmm still find that odd. Like said, 3700 MHz 18-22-22-42 is actually not bad at all on Intel. If that works rock solid, keep it. Notice that your other choice of Gskill Ripjaws V (F4-3600C18D-64GVK, also based on Hynix memory) was 3600 18-22-22-42, so....

That's odd for sure. Anyways, try more relaxed timings if you wish to keep 4000 MHz RAM speed, for example 20-24-24-44 (@ 1.40v or 1.41v) as a baseline, still plenty good for 4000 DDR4. That said, 3700 MHz 18-22-22-42 is actually not bad at all on Intel. If that works rock solid, keep it I'd say.

That motherboard seems overly sensiitive... or the RAM. One of the two for sure. 19-19-19-43 is too tight for anything over 3600Mhz on that Hynix memory, even at 1.41v, I think. Just curious, if you load XMP on BIOS, and then put all "Dram advanced settings" at Auto (exception to "Command Rate" , always 2N), then Save and Exit..... ....what does the motherboard choose for timings on next reboot (if it doesn't crash, that is) ?

Please don't disable Hyper-Threading! You'll want that ENABLED on that i3 12100F, that's for sure! There are no "E-Cores" (Economy Cores) on that particular Intel 12th gen processor, they're all "P-Cores" there (Performance Cores, i.e, the good kind!). You'll want that CPU's 4-Cores assisted by its 8-Threads (Hyper-Threading) at all times, including for DCS. Regarding your BIOS settings, it seems good to me (looks like already "optimized defaults"?) but then I confess to only having read it "in the diagonal".... I did notice the "Command Rate" in the "Dram advanced settings" set at "Auto".... As I've said before, IMO it should always be set manually at "2N" (same thing as 2T) for best stability (even more so if with manual settings in place) but that's your call. So, anyways, I see the 3600 17-19-19-39 RAM new settings in place there, does it work nicely? If yes, how does it compare to your previous RAM working settings?

Yep, I am... its insatiable apetite... every DCS user is an S&M fetishist with it.

Nah, DCS itself doesn't need a pagefile IF you have a lot of RAM memory (64GB or more). DCS does need it if you're stuck with 32GB or less of RAM. Because it can use all that RAM (on busy missions and in MP) and, once it's fully eaten, it'll look out for more elsewhere to devour as emergency resort... that's right, on the pagefile (and yes, it's much much slower!). That's why it's recommended to set the pagefile at 32GB (32768MB) if you have 32GB of RAM memory, and always in the fastest drive.

Some old programs need a minimum of (at least) 80MB+ pagefile, otherwise they won't start or crash. The problem is, at this point there is no list of which ones need it, and which ones don't. So instead of going into that bad lottery (by not using any pagefile), if you have a lot of of RAM memory (64GB or more), and don't want to use an unnecessary big pagefile, then just create a small one for that effect only. For example, I use a pagefile manually set at 1024MB min and 2048MB max.

Gskill Ripjaws V 4000Mhz CL18 (F4-4000C18D-64GVK) is most likely Hynix CJR or Micron Rev.E. There is no Samsung B-die at 4000Mhz 18-22-22-42 - those would be instead 18-18-18-38 or 18-19-19-39 (XMP). That memory besides the 4000 18-22-22-42 (XMP) @1.40v, should also be able to do 3600 17-19-19-39 (maybe @1.35v). You can also try 4000 20-24-24-44 (@1.40v) if you want the speed of that kit. Always use the "Command Rate" at 2T (for better stability). And do change the DRAM GEAR MODE to "Gear 1", as explained above, for any memory at or below 4000Mhz on Intel 12th and 13th gen.

In this particular case it's not the fault of memory XMP. Gskill Ripjaws V are among the most reliable, stable and thoroughly factory tested kits on the DDR4 market (and noticed especially on Intel systems), regardless of the speed/latency/capacity kit. I'm inclined to think it's down to the particular B660M motherboard, and here it's where some small adjustments can make all the difference. Either that, or faulty RAM kit (which would be very unfortunate, but not unique) - if it's that, then perhaps swapping it for the 3600Mhz kit instead should be considered. Unfortunately, yeah.... it can be a bit finicky and require testing, and a lot of patience.

Don't use MSI "Memory Ty it!", as it usually makes things far worse (uses generic algorithms instead of "tried and true" in-depth values). XMP loaded and DRAM Gear Mode at "Gear 1" should work on your Intel 12th gen, but it may require some tweaking on fast ram kits like yours. On the memory timings (if it's on Auto, change to Manual) change the "Command Rate" from 1T to 2T, leaving the rest of timing settings as they already are (which is by XMP profile, if that's loaded). This will relax it a bit in that particular "Command Rate" setting, making it more stable at all times, without affecting performance. DRAM Voltage maybe too relaxed (too short) for 4000Mhz, raise it by 0.025v increments on each try (if it fails). For example, if it's set at 1.400v, manually change it to 1.425v first, if it fails then 1.450v (but avoid more than this, for upper safe limit!). Save and Exit, try again.

That's very odd. Make sure that XMP is set in the BIOS. If it's not, load the XMP profile. After that, make sure (still in BIOS) that the DRAM GEAR MODE is set to "Gear 1". If it's not, manually change it. For example, aside the guy's rambling, you can see how it's done in this video (one among many other out there):

Check the Thermalright Peerless Assassin 120 (regular or SE variant), it goes for just 40 bucks/euros. http://www.thermalright.com/product-category/products/heatsink/dual-towers/peerless-assassin-series/

Same here, OC'ed 10700K (10th gen Comet Lake) still doing great. I'll wait for something (yet to be released) that is trully worth the hefty investment on CPU + Cooler + Motherboard + DDR5 RAM (64GB or 128GB kit). Not even that, as upcoming 14th gen CPUs are just a refresh of current 13th gen Raptor Lake, and Meteor Lake was cancelled for desktop. The 14th gen later this year will have similar IPC (rumoured only 5% IPC gain, at most) and just more "E" cores (which are weaker and not meant for gaming) than current ones. What is definitely worth waiting for is the 15th gen Arrow Lake ("next-gen", multi-tile processor).... Rumoured a 20% increase in IPC over the cancelled Meteor Lake, and a 45% increase in IPC over 12th gen Alder Lake. Which then means a massive ~70% increase in IPC over 10th gen Comet Lake. (now that's an upgrade!) https://www.techspot.com/news/97549-intel-15th-gen-arrow-lake-cpus-how-much.html https://www.notebookcheck.net/Detailed-Intel-15th-gen-Arrow-Lake-leak-suggests-massive-20-IPC-uplift-over-Meteor-Lake-alongside-large-efficiency-improvements.691713.0.html But then, that's a (late?) 2024 release, which also means the following AMD Ryzen (and following X3D?) may also be out too, to compete.

+1

Ditto. This, a million times. We all love detail but, there comes a point that it's unnecessary, it becomes superfluous. * looks at the silly panels details on F15E and BS3... "why??" *

It's an upgrade, but the performance lift will be small considering the investment. Try overclocking that 11600K to 4.9Ghz (all core OC), for which 1.325v Vcore should be enough. OC'ing over 4.9Ghz will require too much voltage to make it stable, so I wouldn't recommend going over that. Plenty of guides in the internet on how to overclock the 11th gen CPUs.

Brilliant mod. Congrats and well done to the author! This "upgrades" all of the FC3 aircraft for the user, makes for a nice and simple middle-ground between their basic (non clickable) default cockpits and the full fidelity (fully clickable) cockpits of far more advanced modules. Thoroughly recommended. Ditto!